IMPACT OF CLIMATE AND LAND USE LAND COVER CHANGE ON

STREAMFLOW: A CASE STUDY OF YADOT RIVER WATERSHED,

GENALE DAWA BASIN, ETHIOPIA.

MSc. THESIS

ABAY MUSTEFA ABDULE

HAWASSA UNIVERSITY, HAWASSA, ETHIOPIA

OCTOBER, 2021
IMPACT OF CLIMATE AND LAND USE LAND COVER CHANGE ON
STREAMFLOW: A CASE STUDY OF YADOT RIVER WATERSHED
GENALE DAWA BASIN, ETHIOPIA.

ABAY MUSTEFA ABDULE

A THESIS SUBMITTED TO THE DEPARTMENT OF WATER

RESOURCES AND IRRIGATION ENGINEERING, INSTITUTE OF
TECHNOLOGY, SCHOOL OF GRADUATE STUDIES
HAWASSA UNIVERSITY
HAWASSA, ETHIOPIA

IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE

DEGREE OF MASTER OF SCIENCE IN WATER RESOURCE
ENGINEERING AND MANAGEMENT

OCTOBER, 2021
 SCHOOL OF GRADUATE STUDIES

HAWASSA UNIVERSITY

ADVISORS’ APPROVAL SHEET

(Submission Sheet -1)

This is to certify that the thesis entitled “Impact of Climate and Land Use Land Cover
Change on Stream flow (A Case Study of Yadot River Watershed Genale Dawa Basin,
Ethiopia)” submitted in partial fulfillment of the requirement for Degree of Master of Science
in Water Resource Engineering and Management, the graduate program of the department
/School of Biosystems and Water Resources Engineering has been carried out Abay Mustefa
ID.NO.WREMK/001/08, under my/our supervision. Therefore, I/we recommend that the
student has fulfilled the requirement and hence hereby can submit the thesis to the department.

Abraham Woldemichael (PhD) _______________________ ____________

Major advisor Signature Date

Alemayehu Muluneh(PhD) _______________________ ____________

Co-advisor Signature Date
 SCHOOL OF GRADUATE STUDIES

HAWASSA UNIVERSITY

EXAMINERS’ APPROVAL SHEET-1

(Submission Sheet -2)

We, the undersigned members of the Examining Board of the Final MSc Open Defense by
Abay Mustefa have read and evaluated his /her thesis entitled “Impact of Climate and
Land Use Land Cover Change on Stream flow (A Case Study of Yadot River Watershed
Genale Dawa Basin, Ethiopia)” and examined the candidate.

This is, therefore, to certify that the thesis has been accepted in partial fulfillment of
requirement for the degree of Master of Science in Water Resource Engineering and
Management.

Abraham Woldemichael (PhD) _________________ _________________

Major Advisor Signature Date
Alemayehu Muluneh (PhD) _________________ ________________
Co-advisor Signature Date
Tewodros T (PhD) _________________ _________________
Chair man Signature Date
Mulugeta D. (PhD) _________________ _________________
Name of Internal Examiner Signature Date
Habtamu H. (PhD) _________________ ______________
Name of External Examiner Signature Date
________________________ _________________ _________________
SGS Approval Signature Date

Final approval and acceptance of the thesis is contingent upon the submission of the final
copy of the thesis to the School of Graduate Studies (SGC) through the Department
/School of graduate committee (DGC/SGS) of the candidate's department.

Stamp of SGS Date: _________________

 Declaration
I hereby declare that this MS thesis is my original work and has not been presented for a
degree in any other university, and all sources of material used for this thesis / dissertation
have been duly acknowledged.

Name: _Abay Mustefa

Signature: _________________

Date of Submission: ____________________

Email: [email protected]
 DEDICATION
This work is dedicated to my lovely wife Fanu Amare, for nursing me with affection, love and
for her dedication in all success of my life.
 ACKNOWLEDGEMENT

I would like to express my sincere gratitude to my major advisor Dr. Abraham W/Michael and
Co-Advisor Dr. Alemayehu Muluneh for their patient, continuous guidance, encouragement
and motivation from the start of proposal writing to the completion thesis work. I won’t to
really thank Dr. Wakjira Takala Dibaba for helping and guiding me to carry out my Msc
thesis.

I extend my appreciation and special thanks to Ministry of Agriculture (MoA) for scholar and
sponsoring me with financial support in order to finish thesis work. Special thanks also go to
Agarfa Agricultural Technical Vocation Education and Training College (ATVET) for
providing all other facilities whenever required during the study.

My acknowledgments extended to the National Meteorological Agency of Ethiopia (NMA)

and Ministry of Water, Irrigation and Energy (MoWIE) particularly hydrology Department for
providing me relevant data including hydro-meteorological and soil shape file required for my
study free of payment.

Lastly, but not least, I would like heart-felt thanks to my wife Fanu Amare , my baby
Abenezer Abay and to all my family for their consistent love, care and encouragement which
served me as a sources of strength throughout the study period.

Above all, I would like to sincere thank the Almighty God, who made it possible, to begin and
finish this work successfully. This day what God have done for me is really beyond what I can
imagine

I
TABLE CONTENTS
ACKNOWLEDGEMENT............................................................................................................I
LIST OF TABLE........................................................................................................................V
LIST OF FIGURE.....................................................................................................................VI
LIST OF ABBREVIATION AND ACRONYMS....................................................................IX
ABSTRACT................................................................................................................................X
1. INTRODUCTION...................................................................................................................1
1.1. Background of the Study...................................................................................................1
1.2. Statement of the Problem..................................................................................................3
1.3. Objectives..........................................................................................................................4
1.3.1.General objective.........................................................................................................4
1.3.2. Specific objectives......................................................................................................4
1.4. Research Questions...........................................................................................................4
1.5. Significance of the Study..................................................................................................4
1.6. Scope of the Study............................................................................................................5
2. LITERATURE REVIEW........................................................................................................6
2.1. The Concept and Definition of Land Use Land Cover.....................................................6
2.2. Land Use Land Cover Changes in Ethiopia......................................................................6
2.3. Land Use Land Cover Change Impact on Stream Flow...................................................8
2.4. Land Use Land Cover Change Prediction.........................................................................9
2.5. Application of Remote Sensing for Land Use Land Cover Change...............................11
2.6. Concept and Definition of Climate Change....................................................................12
2.7. Climate Change in Ethiopia............................................................................................13
2.8. Climate Change Impacts on Stream Flow.......................................................................14
2.9. Projection of Future Climate...........................................................................................16
2.9.1. Global Climate Model (GCM).................................................................................16
2.9.2. Downscaling Techniques (Methods).......................................................................17
2.10. Climate Scenarios.........................................................................................................18
2.10.1. CORDEX-Africa Domain......................................................................................19
2.10.2. Regional Climate Model (RCM)............................................................................20
3.2.1.1. Comparisons of SRES and RCP climate scenarios...........................................20

II
 3.2.1.2. Representative Concentration Pathways (RCP)................................................21
2.11. Hydrological Model......................................................................................................23
2.11.1. Type of Hydrological Model..................................................................................24
2.11.2. Application of SWAT Model.................................................................................25
3. MATERIALS AND METHODS..........................................................................................27
3.1. Description of Study Area...............................................................................................27
3.1.1. Location....................................................................................................................27
3.1.2. Climate.....................................................................................................................28
3.1.3. Land Use Land Cover...............................................................................................30
3.1.4. Soil............................................................................................................................30
3.2. Data Collection Methods and Procedures.......................................................................31
3.2.1. Data Collection............................................................................................................31
3.2.1.1. Digital Elevation Model....................................................................................31
3.2.1.2. Land Use Land Cover Data...............................................................................32
3.2.1.3. Soil Map Data....................................................................................................33
3.2.1.4. Hydro – Meteorological Data............................................................................33
3.2.1.5. Checking Hydro-Meteorological Data quality..................................................34
3.2.2. Land Use Land Cover Change Detection.................................................................37
3.2.2.1. Image Processing and Date of Acquisition........................................................37
3.2.2.2. Image Classification..........................................................................................37
3.2.2.3. Accuracy Assessment........................................................................................39
3.2.3. Land Use Land Cover Change Detection Analysis..................................................40
3.2.4. Land Use Land Cover Prediction and Model Validation.........................................40
3.2.4.1. Land Use Land Cover prediction.......................................................................40
3.2.4.2. Validating LULC Prediction Model..................................................................41
3.2.5. Near-term and Mid-term Climate Projection...........................................................42
3.2.5.1. GCMs and RCMs used for this study................................................................42
3.2.5.2. Bias Correction of Future Climate Data............................................................43
3.2.6. Impact of Climate and LULC Change on Stream flow............................................45
3.2.6.1. Climate Change Impact Simulation Using SWAT Model................................45
3.2.6.2. Description of SWAT Model.............................................................................45
3.2.6.3. Data Input for SWAT Model.............................................................................47
3.2.7. SWAT Model Setup.................................................................................................47
III
 3.2.7.1. Watershed Delineation.......................................................................................47
3.2.7.2. Hydrological Response Unit (HRU)..................................................................48
3.2.7.3. Sensitivity Analysis...........................................................................................48
3.2.7.4. Model Calibration..............................................................................................49
3.2.7.5. Model Validation...............................................................................................50
3.2.7.6. Model Performances Evaluation........................................................................50
4.RESULTS AND DISCUSSION.............................................................................................52
4.1. Land Use Land Cover Classification Accuracy Assessment...................................52
4.2. Land Use Land Cover Change Analysis..................................................................54
4.3. LULC Change Transition Probability Matrix between 1985 to 2015.....................57
4.4. Future Land Use Land Cover Change prediction....................................................59
4.4.1.Validation of CA-Markov Model..............................................................................59
4.4.2. Predicted future Land Use Land Cover....................................................................60
4.5. LULC Transition Probability Matrix between 2015 to 2055..........................................63
4.6. Future Climate Projection...............................................................................................67
4.6.1. Precipitation Projection under RCP4.5 and RCP 8.5 scenarios...............................67
4.6.2. Temperature Projection under RCP4.5 and RCP 8.5 Scenarios...............................71
4.7. Calibration and Validation of SWAT Model..................................................................75
4.7.1. Sensitivity Analysis..................................................................................................75
4.7.2. Calibration of SWAT Model....................................................................................76
4.7.3. Validation of Model.................................................................................................77
4.8. Independent and Combined effect of Climate and LULCC on Stream Flow.................79
4.8.1. Past Land Use Land Cover Change Impact on Stream flow....................................79
4.8.2. Future Land Use Land Cover Change Impact on Stream flow................................81
4.8.3. Impact of Climate Change on Stream flow..............................................................82
4.8.4. Combined Impact of Climate and LULC Change on Stream flow..........................85
5. SUMMARY CONCLUSIONS AND RECOMMENDATION.............................................90
5.1. Summary and Conclusions..............................................................................................90
5.2. Recommendation............................................................................................................92
6. REFERENCE........................................................................................................................93
7. APPENDICES.....................................................................................................................110

IV
 LIST OF TABLES

Table 3.1: Summary of Homogeneity test climate data with 5% (α = 0.05) significance level 36
Table 3.2: Land sat images used for LULC change analyses and their characteristics.............37
Table 3.3: Summaries of GCM and RCM climate model used in the study.............................43
Table 3.4: Performance ratings of recommended statistics for monthly stream flow...............51
Table 4.1: Confusion Matrix for LULC Classification 1985....................................................52
Tables 4.2: Confusion Matrix for LU/LC Classification of 2000.............................................53
Table 4.3: Confusion Matrix for LU/LC Classification of 2015...............................................53
Table 4.4: The area coverage of LULC, percent, and rate of changes in the Ydaot watershed
between 1985, 2000 and 2015...................................................................................................55
Table 4.5: Transition Area Matrix (ha) between 1985-2000 and 200-2015 period in Yadot
watershed...................................................................................................................................58
Table 4.6: Comparison of actual and projected LULC 2015....................................................59
Table 4.7: The k-index values of the simulated LULC map of 2015........................................59
Table 4.8: The area coverage of LULC, percent, and rate of changes in the Ydaot watershed
between 2015, 2035 and 2055...................................................................................................62
Table 4.9: Transition Area Matrix (ha) between 2015 - 2035 and 2035 - 2055 period in Yadot
Watershed..................................................................................................................................65
Table 4.10: Sensitive Flow Parameters, their rank and fitted value..........................................76
Table 4.11: Model Performance Evaluation Statistics for calibration and validation...............78
Table 4.12: Mean Annual and Seasonal changes of stream flow results for past LULC change.
...................................................................................................................................................79
Table 4.13: Mean Annual and Seasonal change of stream flow results for future LULC
change........................................................................................................................................81
Table 4.14: Mean annual and seasonal change of flow under RCP4.5 and RCP8.5 scenarios for
climate change only from baseline period.................................................................................83
Table 4.15: Mean Annual and Seasonal change of flow under RCP4.5 and RCP8.5 scenarios
for combined climate and LULC change from baseline period.................................................86

LIST OF FIGURES
V
Figure 3.1: Location map of Yadot Watershed.........................................................................27
Figure 3.2: Mean Monthly Rainfall Distributions of the Stations (1985-2015)........................29
Figure 3.3: Mean areal rainfalls of the watershed from (1985-2015)........................................29
Figure 3.4: Average Monthly Max, Min and Mean Temp in the period of (1985-2015)..........30
Figure 3.6: Digital Elevation Model of Yadot watershed..........................................................32
Figure 3.7: Double Mass Curves for the Stations......................................................................35
Figure 3.8: Homogeneity tests of annual rainfall data of individual rain gauge stations..........36
Figure 3.9: Pettitt's Homogeneity test of Annual Stream flow..................................................36
Figure 4.1: LULC Map of Yadot Watershed in the period 1985, 2000, and 2015....................56
Figure 4.2: Gain, Loss and Persistence area of LULC class in (1985 -2000) and (2000-2015)58
Figure 4.3: Simulated and observed LULC Map of 2015........................................................60
Figure 4.4: Predicted LULC Map of the year 2035 and 2055...................................................61
Figure 4.5: Gain, Loss and Persistence area of LULC class in (2015 -2035) and (2035-2055)66
Figure 4.6: Mean monthly precipitation under RCP4.5 for near and mid future compared to
baseline period...........................................................................................................................67
Figure 4.7: Percentage changes of monthly rainfall for near-term and midterm under RCP 4.5
scenarios from baseline period..................................................................................................68
Figure 4.8: Mean monthly precipitation under RCP8.5 for near and mid future compared to
baseline period...........................................................................................................................69
Figure 4.9: Percentage changes of monthly rainfall for near-term and midterm under RCP 8.5
scenarios from baseline period..................................................................................................70
Figure 4.10: Mean monthly maximum and minimum temperatures of near and mid future
period with respect to baseline period under RCP 4.5 scenarios...............................................71
Figure 4.11: Mean monthly and seasonal changes in maximum and minimum temperature (°C)
for near and midterm from baseline period under RCP 4.5 scenarios.......................................72
Figure 4.12: Mean monthly maximum and minimum Temperature of near-term and midterm
with respect to baseline period under RCP8.5 scenarios...........................................................73
Figure 4.13: mean monthly and seasonal changes in maximum and minimum temperature (°C)
for near and midterm from baseline period under RCP 8.5 scenarios.......................................74
Figure 4.14: Calibration result for average monthly stream flow (1988-2002)........................77

VI
Figure 4.15: Scatter plots of observed and simulated stream flow for Calibration (1988-2002).
...................................................................................................................................................77
Figure 4.16: Validation result for average monthly stream flow (2003-2008)..........................78
Figure 4.17: Scatter plot of observed and simulated stream flow for Validation (2003-2008).78
Figure 4.18: Mean monthly stream flow results for the past three land use land cover............80
Figure 4.19: Mean monthly flow for near and midterm under RCP4.5 and RCP8.5 scenarios
from the baseline period............................................................................................................84
Figure 4.20: Percentage change of flow for near and midterm future under RCP4.5 and
RCP8.5 scenarios compared to baseline period.........................................................................84
Figure 4.21: Mean monthly stream flow for near and midterm under RCP4.5 and RCP8.5
scenarios compared to baseline period......................................................................................87
Figure 4.22: Percentage change of flow for near-term and midterm under RCP4.5 and RCP8.5
scenarios compared to baseline period......................................................................................88

VII
 LIST OF TABLES IN APPENDIX
Tables 1: Mean Monthly Flow for Only LULC under Baseline and future Period.................110
Tables 2: Major soil type of the watershed..............................................................................110
Tables 3: Meteorological data used for SWAT model (1985-2015) for Delo Mena Station. .111
Tables 4: Mean monthly flow for Climate Change only simulation under RCP4.5 and RCP8.5
.................................................................................................................................................112
Tables 5: Mean Monthly Flow for Combined Simulation under RCP4.5 and RCP8.5...........112
Tables 6: Projected maximum and minimum Temperature for near and midterm underRCP4.5
and RCP8.5 scenarios..............................................................................................................113
Tables 7: Projected mean monthly and % of rainfall under RCP4.5 and RCP8.5..................114

VIII
LIST OF ABBREVIATION AND ACRONYMS
AR5 Fifth Assessment Report
CCLM4.8 Climate Limited Area Model Version 4.8
CMIP5 Coupled Model Inter-Comparison Project Phase 5
CORDEX Coordinated Regional Climate Downscaling Experiment
DEM Digital Elevation Model
EC-EARTH EC-EARTH Consortium
ERDAS Earth System Data Analysis System
GIS Geographic Information System
HadGEM2-ES Hadley Global Environment Model 2- Earth System
LULC Land Use Land Cover
MOWIE Ministry of Water, Irrigation and Electricity
MPI-ESM-LR Max Planck Institute for meteorology (MPI-M)
NSE Nash and Sutcliffe Simulation Efficiency
RACMO22T KNMI, Regional Atmospheric Climate Model, version 22
IPCC Inter Government Panel on Climate Change
RCA4 Rossby Centre Regional Atmospheric Model Version 4
RCM Regional Climate Model
RCP Representative Concentration Pathways
RMSE Root Mean Square Error

SRES Special Report on Emission Scenarios

SWAT Soil and Water Assessment

SWAT-CUP SWAT- Calibration Uncertainty program

IX
 ABSTRACT
Both climate and land use land cover (LULC) change are the main factors that influence
hydrological regimes by altering the magnitude of ground water recharge and river flow.
Thus, for predicting future stream flow both climate and LULC changes projection should be
accounted. In this study, Cellular Automata (CA)-Markov in IDRISI software was used to
predict the future LULC scenarios and the ensemble mean of three regional climate models
(RCMs) in the coordinated regional climate downscaling experiment (CORDEX)-RCM daily
precipitation and temperature for Ethiopia under RCP 4.5 (medium emission scenarios) and
RCP 8.5(higher emission scenarios) were used for the future climate scenarios. Power
transformation and variance scaling method were used to correct bias the RCMs outputs, with
respect to the observed precipitation and temperature. The separate and combined impact of
climate and LULC change on stream flow was analyzed using SWAT hydrological model. The
calibrated and validated for stream flow simulation using SWAT-CUP with a method of
SUFI2.The performance of the model was assessed through calibration and validation process
and resulted R2 = 0.8 and ENS = 0.73 during calibration and R2 = 0.83 and ENS = 0.77
during validation on monthly base simulation. The results of the ensemble mean of the three
RCMs (CCLM4.8, RACMO22T and EC-EARTH) output show parallel precipitation and
temperature increasing trends in the future under RCP4.5 and RCP8.5 scenarios but vary on
monthly basis. The increases in mean annual maximum and minimum temperatures are higher
for higher emission scenarios than medium emission scenarios. The LULC results showed that
both in the past and future period, agricultural and settlement are significantly increased
while forest land and scrub/bush lands continuously declined conversely grass/range lands
and wood land show decline in the past and increased from 2015 to 2035 and again decreased
from 2035 to 2055 in the future period. The past LULC caused an increased mean annual flow
by 1.26%, and wet season flow by 2.68% but dry season flow decreased by 2.22% while the
future LULC 2015 to 2055 will cause mean annual flow increased by 1.19%, and wet season
flow by 2.9% but by decreased for dry season flow by 3.14%. The mean annual flow is
projected to increase under both climate and combined scenarios by 7.63% (8.13%) and
5.76% (6.26%) in the near (2021-2050), while in the midterm (2051 – 2080) flow increased by
5.76% (6.26) and 6.07% (6.72%) at the outlet of the watershed under RCP4.5 and RCP8.5
scenarios, respectively. Generally, results of future stream flow projection indicated that the
combined change of climate and LULC have relatively higher than the climate changed alone.
Such studies enhance better understanding of the various impacts of climate and LULC
change scenarios on stream flow, which can be used for better adaptation and mitigation of
water resources management problem in the watershed by Appling different water and soil
conservation measures.
Key Words: Climate Change, Land use/land cover change, CA-Markov, RCP, RCM

X
XI
 1. INTRODUCTION
1.1. Background of the Study

Climate and land Use land Cover change have significant impacts on the hydrology of the
watershed through changing processes such as precipitation, evapotranspiration and
infiltration, thus affecting the rainfall-runoff processes (Yin et al., 2017) and the amount of
available water resources (Li et al., 2012). Stream flow, as a primary component of the
hydrology and widely believed to be affected mainly by climate and land use land cover
changes (Ning et al., 2016 and Wuletawu et al., 2019).

Changes in climate and land use land cover are the two inseparable linked global
environmental challenges the world faces today (IPCC, 2019). Since the beginning of the
industrial revolution in the mid-eighteenth century, the contribution of human activity to
climate change has increased dramatically, by increasing the concentration of greenhouse
gases (GHGs) in the atmosphere. The rise in temperature and precipitation patterns are
prominent features of change in climate that directly impact almost all other hydrological
responses (Azari et al., 2016). Therefore, understanding and quantifying their respective
influence is of great importance for water resources management and socio-economic
activities as well as policy and planning for sustainable development (Yang et al., 2017).

Water resources are currently under severe pressure because of impacts of climate change and
human activities, which include land-use change, increasing population growth, and economic
development (IPCC, 2013). Climate change due to human activity has led to changes in
precipitation patterns and increased the frequency of heavy precipitation events which have
been significant effects on the ecological system ( Belay et al., 2017) and affects streamflow
through changes in temperature, precipitation, and evaporation (Ahn and Merwade, 2014; Guo
et al., 2019). Land use land cover change can significantly alter canopy interception,
infiltration and evapotranspiration, which may eventually change the runoff volume, peak
flow and flow routing time (Zhang et al., 2017and Umair et al., 2019). However, the relative
impacts of land use land cover and climate change may vary from place to place due to the
rate and extent of changes in climate and land-use. Therefore, it is crucial to the long-term

1
water resource planning and management to better understand the potential impacts of climate
and land use land cover changes on the runoff and stream flow in the basins.

In recent years, many studies examined the combined impact of land use land cover and
climate change on hydrologic processes (Koch et al., 2015; Hyandye et al., 2018; Aboelnour
et al., 2019). The findings from different studies consistently highlight the water balance
components including streamflow, surface runoff, groundwater, and evapotranspiration are
likely to be impacted by future land use and climatic changes. Determining the individual or
combined hydrological consequences of climate and land use land cover changes is a key for
implementing effective measures for adaptation of climate change and understanding the
patterns of water use under different land use land cover policies (Wang et al., 2018; Clerici et
al., 2019; Trolle et al., 2019). For example, some studies have found that the hydrological
processes are impacted more by climate change than land-use change (Dagnenet et al., 2018;
Aboelnour et al. , 2019), and other studies have indicated that the impacts of land-use change
are more significant as compared to the impacts of climate change (M wangi et al. , 2016; Yin
et al. , 2017). On the other hand, enhancing land use land cover will mitigate the negative
effects of climate change (Tekalegn et al., 2018; Mulatu et al., 2019). This suggests that the
combined effect of climate and LULC change on hydrological responses requires site specific
investigation.

There are many effective models that can continuously simulate stream flow, erosion or
nutrient loss from a watershed. From those hydrologic models the Soil and Water Assessment
Tool (SWAT) has been widely applied to understand the relationship between climate and
LULC change on hydrology (Chang et al., 2015, Natkhin et al., 2015). This study aims to
assess the impacts of climate and land use land cover changes on stream flows of Yadot river
watershed using SWAT Hydrological model and downscaled hydro climatic data from
CORDEX-RCMs output under RCP4.5 and RCP8.5 climate scenarios within Genale Dawa
basin of Ethiopia. Integrating remote sensing-based high resolution satellite image of temporal
land use land cover classification with hydrological modeling would significantly improve the
accuracy in simulating the impacts of the LULC change on the hydrologic regime.

2
 1.2. Statement of the Problem

Yadot watershed is one of the watersheds draining in to river basin of Genale Dawa, and it is
under threat due to the growing population and increasing demand of water mainly for
irrigation and other development activities. The watershed is intensively cultivated and there is
high expansion of settlement in the watershed. The watershed is undergoing climate and land
use land cover change due to expansion of agricultural land and unplanned and unrestricted
settlement as a result of rapid population growth, deforestation (fire), improper land use and
rapid immigration which has an impact on hydrologic response of the watershed. Among these
changes in climate and land use land cover have a significant impact on the hydrologic regime
by affecting stream flow and volume of rivers flow and evapotranspiration. Therefore,
understanding watershed hydrology is crucial for effective and sustainable water resource
planning, development and management activities.

Most of the study in the basin focused on the impact of either climate change (Tufa & Sarma
2021; Tesfaye et al., 2019) or land use land cover change (Mesfin, 2018) on stream flow
response. There was a clear general lack of modeling on the combined impact of land use land
cover and climate change on watershed hydrology. In terms of land use land cover and climate
change, major participants that alter a stream flow in a study area are urbanization, agricultural
and deforestation activities. This continuous change in land use land cover and climate has
impacted the water balance of the watershed by changing the magnitude of the components of
stream flow which are surface runoff and ground water flow, which results increasing the
extent of the water management problem. Therefore, a properly managed rivers watershed can
provide fresh drinking water, food, hydropower and recreation. However, there were no
comprehensive studies which projected the future impact of climate and land use land cover
changes on the hydrologic regime of the basin. Therefore this study tried to bridge this
knowledge gap by assessing the impacts of current and future land use land cover and climate
change on stream flow of Yadot Watershed.

3
 1.3. Objectives

1.3.1. General objective

The general objective of this study is to assess the impact of climate and land use land cover
change on stream flow using SWAT hydrological model and downscaled hydro climatic data
from CORDEX-RCMs output under RCP4.5 and RCP8.5 climate scenarios: A Case Study of
Yadot River Watershed Genale Dawa Basin, Ethiopia.

1.3.2. Specific objectives

 To detect the past and predict the future land use land cover changes in the Yadot
Watershed.
 To assess the projected climate change for the near and midterm in Yadot watershed
 To evaluate the independent and combined effect of climate and land use land cover
changes on stream flow in the Yadot Watershed.

1.4. Research Questions

How are the past and future land-use/land cover changes over time in Yadot Watershed
from 1985 to 2055?
What are the projected changes in precipitation and temperature in the Yadot Watershed in
the near and midterm period?
What is the trend of the past and future land-use/land cover change impact on stream flow?
What is the combined effect of Land-use/land cover and climate change to the stream flow
from near to midterm period in the study area?

1.5. Significance of the Study

The land use land cover and climate change has a significant impact on the natural resources,
of the watershed, socio economic and the future potential impacts of these natural resources
with respect to the changing environment. However, to determine the effects of climate and
land use land cover change on stream flow, it is important to have an understanding of the
climate and land use land cover change patterns and the hydrological processes of the
4
watershed. Therefore, the finding research will strongly assist governments and
decision/policy makers in planning, formulate and implement effective and appropriate
response strategies to minimize the undesirable land use land cover and climate change
impacts and any other development activities in a watershed.

The significance of a research conducted in the proposed study area is sound because it adds
initiatives in watershed management by enhancing the understanding of impacts of various
environment (land use land cover changes) and climate change scenarios which, consequently
indicate a potential for measurable changes in the conditions that control water yields
including rainfall, temperature and stream flow. In generally, this study will be expected to fill
the gaps in incorporating the impact of land use/land cover and climate change on stream flow
particular in Yadot River watershed.

1.6. Scope of the Study

This study is a step to understand the combined impact of climate and land use land cover
change on stream flow response of the watershed by using different historical and future land
use land cover and meteorological data available for climatic analysis in SWAT flow model.
This study is limited to look at climate and land use land cover changes impact on stream flow
in both historical and future period within in a Yadot River watershed Genale Dawa.

5
 2. LITERATURE REVIEW

2.1. The Concept and Definition of Land Use Land Cover

The terms land use land cover often has been confusing and used interchangeably in the
literature and also in daily practice. Thus is it important to define and understand the meaning
of these terms. Land cover change refers to modification of the existing land cover or
complete conversion of land cover to a new cover type (Wu et al., 2013). Land use refers to
the intended use or management of the land cover type by human beings. Thus, land use
involves both the manner in which the biophysical attributes of land are manipulated and
intent underlining that manipulation (the purpose for which the land is used e.g., agriculture,
grazing), which are more subtle changes that affect the character of the land cover without
changing its overall classification (Di Gregorio, 2016).

Both conversion and modifications of land use land cover have important environmental
consequences through their impacts on soil and water, biodiversity and contribute to
watershed degradation (Obahoundje et al., 2017). Generally, knowing of impacts of land use
land cover change on the natural resources like water resources depends on understanding of
the past land use practices, current land use land cover patterns, and projection of future land
use land cover, as affected by population size and distribution, economic development,
technology, and other factors

2.2. Land Use Land Cover Changes in Ethiopia

In Ethiopia, land is used for agriculture, pasture, as forest area, buildings and infrastructure
sites or for recreational purposes. With the Rapid population growth and slow modern
agricultural technological adoption practices in the country increase pressure in deforestation
for more production which means conversion of forest to agricultural land and expansion of
urban settlements. The dynamic nature of land use arising from an increasing population at an
alarming rate in Ethiopia (Haile and Assefa, 2012). Most land use land cover changes in
Ethiopia are caused by human drivers mainly due to population growth (Temesgen et al.,
2014). The anthropogenic activities result in an expansion of agricultural land and
urbanization thereby deforestation.

6
Accordingly, some remnant stands of natural forests are mainly restricted to religious sites,
along rivers and streams, and on peaks of hills where crop cultivation is difficult in the
highlands of Ethiopia (Warra et al., 2013). This is due to the rapid expansion of agriculture
and rural and urban settlement. Land use land cover changes in Ethiopia, particularly in the
highland areas caused by a combination of various factors though it depends on the conditions
of the area (Hassen et al., 2015). Because of population pressures, economic factors and policy
issues, settlements, farmland and degraded lands have been expanding while grasslands and
forest areas have been diminishing largely (Binyam et al., 2015; Hassen et al., 2015). Study by
(Ebrahim and Mohamed, 2018) reported that farmlands and settlement areas were expanded at
the expense of forest area, shrub lands, and grasslands over the study period between 1957 and
2014 in Gelda watershed, Lake Tana Watershed of Ethiopia.

Most of these studies indicate that croplands have expanded at the expanse of natural
vegetation including forests and scrublands; for example (Minichil et al., 2016; Temesgan et
al., 2018; Gudeta, 2016) in Blue Nile basin of Ethiopia. Study on land use and land cover
change in the Bale mountain eco-region of Ethiopia showed that forest cover decreased from
20.8 % in 1985 to 20.6 %, 18.8 % and 17.5 % in 1995, 2005 and 2015 respectively. Grass land
and shrub land also decreased continuously. But farmland and urban settlement increased
continuously. Farm land and urban settlement expansion were found to be major drivers of
land use and land cover change (Sisay et al. 2016). Similarly, (Yesuph and Dagnew, 2019)
reported the consistent increase in farmland and settlement area at the expense of Afro-alpine
and sub-Afro-alpine vegetation areas between 1973 and 2017 in Beshillo Watershed of Blue
Nile Basin, Northeastern Highlands of Ethiopia.

The observed land use and land cover changes were driven by population growth, agriculture
land and rural settlement expansion, and growing demand of forest for extraction of fuel and
construction materials (Birhan and Assefa, 2018). The main reason behind higher number of
land less in Ethiopia as reported by Fasil (2012) is population growth which converted large
areas of grass land or forest land into crop or wood land. In the Gilgel Gibe watershed in the
south western part of Ethiopia agricultural land and urban area increased continuously in 1987,
2001 and 2010 all these years (Wakjira et al., 2016). Especially in the highlands of Ethiopia,
agricultural practices and human settlement have a long history and recently a highland

7
population pressure including depletion of natural resources and unsustainable practices (Leal
et al, 2021). Therefore, rapid increase in deforestation as well as poor practices of managing
farmlands accelerating soil erosion and land degradation in the Ethiopian highlands (Hassen et
al., 2015).

2.3. Land Use Land Cover Change Impact on Stream Flow

Land use land cover changes have major impacts on hydrological processes, such as runoff
and ground water flow (Asmamaw, 2013). It is obvious that land use land cover can affect
both the degree of infiltration and runoff following rainfall events, while the degree of land
cover can affect rates of evaporation. It was reported by (Amare, 2013) that land use land
cover changes affected the stream flow of Gilgel Abbay Watershed, Ethiopia. the study by
(Yitea and Van, 2015) identified that there was an increase of stream flow by 16.26m 3/s during
wet months and decreased by 5.41 m 3/s from 1986 to 2015 reported that the growth of
population and its effect on the land use land cover change has been influencing the hydrology
of the Melka Kuntrie basin, in the Upper Awash River basin by changing the magnitude of
stream flow and groundwater flow. The land use in 2003, which was mostly converted to
agriculture land from forest, grass, or shrub land compared to 1986 land use and the result
shows that increased stream flow in the main rainy season, while the stream flow in dry or
small rainy season indicted inconsistency from month to month. This result is in line with the
finding of (Welde, 2016) which indicated the increment in bare land and agricultural land
resulted in increased annual and seasonal stream flow and sediment yielded in volumes in
Tekeze Dam watershed.

The study by (Abebe, 2020) analysis has shown that the cultivated land has increased from
60.69% to 67.17% and urban land from 2.3% to 3.36% between 1998 and 2016. Whereas the
grassland area has decreased from 11.42% to 5.33%, plantation forest from1.84% to 0.9%, and
bare land from 3.58% to 2.56%. Result in the increase means annual stream flow 150.3 m 3/sec
to 165.6m3/sec for 1998 and 2016 due to land use change in Megech Dam Watershed,
respectively. Rapid population growth increases the crops land by decreases the forest land
which results increase in stream flow because of soil moisture demand (Asmamaw, 2013).
(Gebrie, 2016) concluded that the land use land cover change have a great influence on stream

8
flow especially during wet season than dry season. Increasing land use conversion (especially
for urbanization, deforestation, grassland depletion) can potentially lead to an increase in
stream flow and flood frequency (Rajib and Merwade, 2017).

The conversion of the land surface from native cover to managed cropland has an effect on the
evapotranspiration, infiltration and overland runoff characteristics of a watershed. The
increased removal of native vegetation and soil compaction decreases soil infiltration capacity.
Hence, this leads to an increase in stream flow. As the watershed becomes more developed, it
also becomes more hydrological active, changing the flood volume and runoff components as
well as the origin of stream flow (Tesfa, 2015). The decrease of forest land and grass land was
accompanied by the increase in agricultural and built up areas and this change in land use land
cover increased surface runoff during wet seasons and reduced base flow during the dry
seasons (Getachew and Melesse, 2012).

During storm events, greater surface runoff can exceed the flow carrying capacity of the
stream within the watershed which may increase the risk of potential flooding. Understanding
how land use land cover change influence stream flow will enable planners to formulate
policies to decrease the effect of future land use change on stream flow. Surface runoff is
extremely limited under grass or forest vegetation compared with agricultural land. Therefore
different types of land use land cover have an impact to increase or decrees stream flow.
Quantifying the impacts of land use land cover change practices on the hydrological response
of a watershed has been an area of interest for the hydrologists in recent years as this
information could serve as a basis for developing sound watershed management interventions
(Abdi et al., 2014).

2.4. Land Use Land Cover Change Prediction

Markov Chain Model was presented by a Russian mathematician named Andrei A. Markov
in 1970. Markov chain analysis is a convenient tool for modeling land use land cover (LULC)
change when changes and processing in the land use land cover (LULC) are difficult to
describe. Markov is a stochastic model which requires pairs of LULC images (Clark labs,
2012; Eastman, 2012). Markov chains are stochastic processes (Halmy et al., 2015) and the
matrices to show changes between land use categories (based on the basic core principle of

9
continuation of historical development) (Koomen and Borsboom-van Beurden, 2011) and are
often used in modeling and simulation changes and trends of LULC (Halmy et al., 2015;
Mishra and Rai, 2016; Parsa et al., 2016). The homogeneous Markov model for prediction of
land use changes can be mathematically presented as follows (Subedi et al., 2013):

L(t +1) =P ij∗L(t )−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−¿1

{
P 11 p 12 … … p 1n
P 21 p 22 … … P2 n
Pij = … … … … … … … … … … … … … … … … … … … … … … . … … … … … … .. … … … … … .2
……………………… …
Pn 1 Pn 2 … … .. pnn

¿Where L(t) and L(t+1) represent land use status at time t and t + 1 respectively. Pij is the
transition probability matrix in a state; Pn is the number of LULC categories

Cellular Automat (CA) is a spatial dynamic model which predicts transitions among any
number of LULC categories (Li et al., 2015). The CA-Markov analysis was used to test run a
pair of land cover images and outputs a transition probability matrix and a transition areas
matrix. The transition probability matrix explains the probability that each land cover category
changed to every other category. The transition areas matrix is the number of pixels that are
expected to change from each land cover type to every other land cover type over the specified
number of time units. CA consists of a grid or a raster space, a set of states characterizing the
grid cells and a definition for the neighborhood arrangement of cells, a set of transition rules
determine the state transitions for each of the cells as a function of the position of neighboring
cells and a sequence of discrete time steps then updates composition and configuration of all
the cell simultaneously (Arsanjani et al., 2013; Mishra and Rai, 2016; Parsa et al., 2016).

The basic principle of CA is that the land use changes for any location (cells) can be explained
by the current state and changes in neighboring cells (Koomen and Borsboom-van Beurden,
2011). Mathematically, CA model (Al-sharif and Pradhan, 2013) is expressed as

10
S(t ,t +1) =f (S ( t ) , N )---------------------------------------------------------------------------------------3

Cellular Automat –Markov (CA-Markov) is a robust model which integrates the ability of
Markov and CA models (Li et al., 2015). Cellular automata added into a Markov model lead
to probable spatial transitions occurring in particular area over a period time (Subedi et al.,
2013). In other words, the quantity of changes from the Markov Chain model then is made
geo-referred and spatial through cellular automata (Mishra and Rai, 2016). The CA-Markov
model uses Markov Chain analysis outputs, particularly the transition area file, to apply a
contiguity filter to enable the development of other land use characteristics from two LULC
into ( combined one LULC) one a later time period (Parsa et al., 2016). CA-Markov model is
considered a robust approach because of the quantitative estimation and the spatial and
temporal dynamic it has for modeling the LULC dynamic (Mishra and Rai, 2016; Parsa et al.,
2016). In this study, IDRISI Selva v.17 is used to predict the future LULC of study area on
CA-Markov model.

2.5. Application of Remote Sensing for Land Use Land Cover Change

Remote Sensing is the science and art of obtaining information about an object, area or
phenomenon through an analysis of the data acquired by a device which is not in contact with
the object, area or phenomenon under investigation (Reddy, 2008). Remote Sensing (RS),
integrated with Geographic Information System (GIS) provides an effective tool for analysis
of land use and land cover changes (Afera et al., 2018). It is observed that, Remote Sensing
and GIS is the most modern technology widely used in natural resource management and
monitoring and is a helpful tool in detecting and analyzing spatiotemporal land use/land cover
dynamics and evaluation of land use and land cover changes at watershed levels (Ermias et.al.,
2013). Some of the application of remote sensing technology in mapping and studying of the
land use and land cover changes are; map and classify the land use and land cover, assess the
spatial arrangement of land use and land cover, allow analysis of time-series images used to
analyze landscape history, report and analyze results of inventories including inputs to
Geographic Information System (GIS), provide a basis for model building.

(Andualem et al. 2018) applied remote sensing to detect the land use of Upper Rib watershed.
Their result implies the huge reduction in grassland and increment in cropland (13.78%) for
11
over an 11year period. This was due to the increased population with great interests in
agricultural land. Asirat, (2018) also applied Remote Sensing and GIS for evaluating land use
land cover dynamics in Upper Blue Nile and reported the increment in cropland and grassland,
while a reduction in forest and scrubland. The changes were due to an increment in the human
and livestock population in the area. To monitor the rapid changes of land cover, to classify
the types of land cover, and to obtain timely land cover information, multi temporal remotely
sensed images are considered effective data sources. Remote sensing and GIS technique have
been used extensively to provide accurate and timely information describing the extent of
LULC over time. Change detection analysis, employing both GIS and remotely sensed data,
has been used to assess forest depletion (Soni et al., 2015; Afera et al., 2018) in the Yadot
watershed.

2.6. Concept and Definition of Climate Change

A simple definition climate is the average weather condition of a region over a long period of
time. According to IPCC (2013) climate change can be defined as a change in the state of the
climate that can be identified by their change and variability of its properties and that persists
for an extended period, typically decades or longer. Also (IPCC, 2013) defines climate change
as a change in the state of the climate that can be identified by changes in the mean and/or the
variability of its properties and persist for an extended period typically a decade or even longer
either due to anthropogenic or natural forcing. Climate change is the most serious problem that
the whole world is facing today. It is now widely accepted that climate change is already
happening and further change is inevitable;

The total increase of 0.780C in global temperature from the second half of the 19 century to 1 st
decade of 20th century (IPCC 2013) There is also change in precipitation over the global land
areas, particularly Northern Hemisphere has visible increase in precipitation after 1951
(Hartmann et al. 2013). Between the start and end of 20 th century, the global surface
temperature was increased roughly by 0.74 0C (Shongwe et al., 2015). The average global land
and ocean surface temperature data calculated shows a warming of 0.85 0C in the period
of1880 to 2012 (Shongwe et al., 2015). The IPCC’s latest findings, global average
temperatures will probably raise a further 1.1 to 6.4ºC this century, depending on the extent of

12
continued greenhouse gas emissions. Concentration of carbon dioxide (CO 2), methane (CH4)
and nitrous oxide (N2O) increase since 1750 by 40%, 150% and 20%, respectively (IPPC,
2014). The IPCC 5th Assessment Report (AR5) concluded that most of the observed increase
in global average temperature since the mid-20 th century is very likely due to the observed
increase in anthropogenic greenhouse gas concentrations.

2.7. Climate Change in Ethiopia

Like most of Africa countries, Ethiopia has become warmer over the past century and human
induced climate change will bring further warming over the next century at unprecedented
rates, (Charles et al., 2011). For the past four decades, the average annual temperature in
Ethiopia has been increasing by 0.37°C every ten years, which is slightly lower than the
average global temperature rising (Osima et al., 2018: Emerta and Aragie, 2013). According
to (Emerta and Aragie, 2013) the greater part of the temperature rise was observed during the
second half of the 1990’s and temperature rise is more pronounced in the dry and hot spots of
the country, which are located in the northern, northeastern, and eastern parts of the country.

According to Kassie et al., (2014), Ethiopia’s mean annual temperature is showing a

significant warming trend leading to increasing rates of evapotranspiration and crop water
requirements, further adding to the already frequent water stress of crops. Future projections
show that the mean maximum temperature will increase from 2 to 2.3°C until 2030 and from
2.2 to 2.7 °C until 2050, while the mean minimum temperature will rise between 0.8 and
0.9°C until 2030 and between 1.4 and 1.7°C until 2050, all in conjunction with a surge of hot
days and nights and a decrease of cold days and nights (Gebre et al., 2015). According to
many studies, the precipitation projection in the next future shows an increase and decrease.
Climate change is one of the most investigated issues of Ethiopian climate since its rain-fed
agriculture is largely dependent on the amount and regular onset of seasonal precipitation
(Gebre et al., 2013).

Climate change affects various dimensions of Ethiopia’s economy including hydro-power

generation; the country’s health, education and industry sectors; and public spending on
productive investments. However, estimating the impact of climate variability on agriculture is
more straightforward and easy to quantify (Emerta and Aragie, 2013). Ethiopia is a country
13
where about 80% of the population is engaged in the agricultural sector and the main source of
income for rural communities. With growing populations, industrialization, climate change
and its variability the situation becomes more and more tense (Goitom, 2014). The IPCC 2014
findings indicate that developing countries, such as Ethiopia, will be more vulnerable to
climate change. It may have far reaching implications to Ethiopia for various reasons, mainly
as its economy largely depends on agriculture. Large share of Ethiopia’s population is very
vulnerable to climate change and preparing adaptation options as part of the entire program is
very crucial in particular to its inter-annual variability (Bekele et al., 2014; Headey et al.,
2014). Ethiopia’s economy depends heavily on agriculture and faces increasing population
growth. Ethiopia’s agricultural sector contributes 47% of the country’s gross national product
and more than 80% of its exports. Most of the global climate models project an increase in
precipitation in both the dry and wet seasons.

2.8. Climate Change Impacts on Stream Flow

The changes in stream flow characteristics resulting from climate change depend on individual
watershed aspects. The impact of climate change on stream flows, soil moisture, ground water
and other hydrological parameters essentially involves taking projections of climatic variables
(e.g., precipitation, temperature, humidity, mean sea level pressure) at a global scale.
According to (Anwar et al., 2016) Stream flow projections for future time periods showed that
mean annual stream flow may increase by 7.1, 9.7, and 10.1 % at 2020s, 2050s, and 2080s,
respectively, from the baseline period for A2 scenario, whereas for B2 scenario, it will be
expected to increase by 6.8, 7.9, and 6.4 % for 2020s, 2040s, and 2080s, respectively in the
upper Gilgel Abay Catchment, Blue Nile Basin. (Megersa et al., 2021) study in Upper Wabe
Bridge watershed found in the Wabe Shebele basin reveals that annual stream flow shows
increase trends under RCP 4.5 and RCP 8.5 scenarios in both (2041–2070 and 2071–2099)
projection periods. The mean seasonally stream flow decreases in the Belg (short rainy) season
by −10.91% and increases in the Kiremt (rainy) season by 17.4%. Hence, the increment trend
in stream flow during the Kiremt season will have the utmost significance for the requirement
of storage and rainwater harvesting structures for irrigation development during dry season.

14
(Andargachew and Fantahun 2017) in Gumara catchment, in the Lake Tana Basin-Upper Blue
Nile Basin showed that Seasonal and annual flow volume increases in all future time horizons
as compared to the base period. Seasonally, the maximum increment was shown in the major
parts of the rainy season (Kiremit) and small rainy season (Belg) in which the flow volume
increases by 144.65% for A2a scenario and 101.58% for B2a scenario. Moreover, the annual
increment showed systematic trends and the increment reaches up to 17.65% for both
scenarios at the end of 21st century. (Tesfaye et al., 2019) Awata River Watershed, Genale
Dawa Basin conclude the average total annual flow at the outlet of the watershed might
increase up to 7.3% for the RCP4.5 scenario and 7.0% for the RCP8.5 scenario for the 2018-
2047 periods and for 2048-2077 periods it might increase up to 7.7% for RCP4.5 scenario and
7.9% for the RCP8.5 scenario. In conclusion increase in average total annual, seasonal and
monthly flow volume is observed for periods which show a corresponding increase in mean
annual, seasonal and monthly precipitation during scenario developments.

Several studies conducted in parts different parts of Africa showed that understanding the
impacts of climate change is vital especially in regions where the social and ecological
systems are highly dependent on climate and water resources. For example, (Aich et al. 2014)
assessed the impact of climate change on stream flow in four large representative river basins
(Niger, Upper Blue Nile, Oubangui and Limpopo basin) and found a statistically significant
increase in stream flow for the period 2070 to 2099 compared to the baseline period 1970 to
1999. Study by (Amisigo, 2018) Results obtained the impact assessment showed that future
mean annual stream flow into the Volta Lake could increase by about 17% and 16% under the
A1B and A2 scenarios, respectively. However, stream flows from a few sub basins in the
Black and White Volta basins are projected to decrease due to climate change. The climate
change is less likely to pose the threat on average water availability. However, temporal
variation in river flows is expected to increase in the future (Devkota and Gyawali, 2015).

According to Liersch et al. (2018) also showed that climate change may decrease stream flow
in the Upper Blue Nile River for the months June and July, but an increase for the period
August to November. For example at the local scale (Tufa and Sarma 2021) evaluated climate
change-induced impact on stream flow and sediment yield at Genale Watershed, Ethiopia
using the SWAT model and statistically downscaled CORDEX-RCM and showed that the

15
increase in discharge in March, April, May, August, September under RCP4.5. However, the
same study (Tufa and Sarma 2021) reported more pronounced increase in the same months
under RCP8.5 scenarios. Average monthly change of stream flow for the RCP4.5 and RCP8.5
was running from -16.47% to 6.58% and -3.6% to 8.27%, respectively, of 2022–2080.

The study (Mesgana et al., 2017) on potential impact of climate change on the streamflow of
four major river basins in Ethiopia: Awash, Baro, Genale, and Tekeze show that the ensemble
mean of 10 GCMs indicate the temperature in those area increase by about 2.3 °C (3.3 °C) in
2050s (2080s), whereas the mean annual precipitation is projected to increase by about 6%
(9%) in 2050s (2080s). This results in about 3% (6%) increase in the projected annual
streamflow in Awash, Baro, and Tekeze rivers whereas the annual streamflow of Genale river
is projected to increase by about 18% (33%) in the 2050s (2080s). However, such projected
increase in the mean annual streamflow due to increasing precipitation over Ethiopia
contradicts the decreasing trends in mean annual precipitation observed in recent decades.
Therefore in Ethiopia suitable planning and sustainable utilization of water resources
development based on climate change impact is very critical

2.9. Projection of Future Climate

2.9.1. Global Climate Model (GCM)

Global Climate Models (GCMs) are the primary tools for understanding the global climate and
its projected change under different forcing scenarios. One of the best tools for simulating
current and future predictions of climate change scenarios is a global climate model (GCM)
(Tang, Xu and Xu, 2012). Many GCMs illustrate global climate in 3D-grid with horizontal
resolution from 2.50 latitudes by 3.750 longitudes for atmospheric component to 1.25 0 latitudes
by 1.250 longitudes for oceanic component. However, the coarse resolution of GCMs
precludes them from capturing the effects of local forcing like terrain effect and land-sea
contrasts that modulate the signal at a finer scale (Rummukainen, 2010).

Global climate models also known as general circulation models (GCMs) have significant
implications for climate at global and regional scales by indicating that rising concentration of
greenhouse gases. According to many research results GCMs are the vital resources used to

16
perform climate change experiments regionally, globally and very fine scale up to point
climate pattern from which climate change scenarios are derived; but they have main
drawbacks because of their course resolution. A significant spatial scale problem exists
between the scale of the GCMs (200-400km) and the scales of interest for impacts and
adaptation studies which are often only tens of kilometres or less (Evans, 2011). Their
resolution is thus quite coarse relative to the scale of exposure units in most impact
assessments. The GCMs simulation skill decreases from climate variables to hydrological
variables while the hydrological importance increases along the same direction (Xu, 1999).
Further (Xu, 1999) has identified three different gaps in using the GCMs for water resources
studies. These are: (i) The spatial and temporal scale mismatches, (ii) The vertical level
mismatches, and (iii) The accuracy mismatch (IPCC-TGICA, 2007).

2.9.2. Downscaling Techniques (Methods)

Downscaling is the process of translating the climate projections from coarse resolution GCMs
to finer resolutions RCMs (Patte et al., 2014). Downscaling is dependent on the ability of
GCMs to successfully project the climate change signal; it is limited to where that signal is
clear. Downscaling process is an agent that establishes relationship between climate variables
on a local scale (predicted) and large-scale atmospheric variables (predictors) (Wilby and
Dawson, 2013). Downscaling used to increase the temporal resolution (e.g. from monthly to
daily values of precipitation and temperature) or the spatial resolution (e.g., from a GCM grid
cell size down to the weather station scale). One of the most important parts of downscaling is
its ability to identify and analyze extreme events (Domínguez et al., 2013). Downscaling
approaches are generally categorized as dynamical, using regional climate models, and
statistical, using empirical relationships. Dynamic and statistical downscaling is used for
developing local or regional scales information from the global climate scenarios applying
global climate model.

Statistical downscaling techniques, which employ a statistical relationship between the large-
scale climatic data and local variations derived from historical data (Chen, Brissette and
Leconte, 2012). Statistical downscaling involves the establishment of empirical relationships
between historical and/or current large-scale atmospheric and local climate variables.
Statistical downscaling is based on statistical relationships linking regional climate variables

17
to large scale atmospheric variables (predictors). Such links are determined during an
observational period, tested with independent data outside this period, and used for computing
future climate projections (Pizzigalli et al., 2012). Advantages of statistical modeling also
include the opportunity to use “ensemble” GCM results. On the downside, the main
disadvantage of this approach is that it requires long historical meteorological weather station
data to construct an appropriate link with large-scale variables.

Dynamical downscaling usually involves a nested or higher spatial resolution of 10-50 km of

regional climate model (RCM) through the inclusion of more realistic topography and land
use/vegetation (Chen, Brissette and Leconte, 2012). In dynamic downscaling systematic errors
will occur due to RCMs, require statistical corrections to provide realistic output. Therefore
dynamical downscaling often includes statistical modeling in the form of bias correction.
Dynamical downscaling utilizing higher resolution regional climate models (RCMs) that
respond to the output of GCMs. The GCM output is provided as boundary conditions, which
are the values at the edges of the spatial domain of the RCM.

The main drawbacks are the requirement of powerful computing capacities and the
dependency on initial and boundary conditions. There is also still a lack of readily available
climate scenario ensembles for most regions in the world, although the number of publically
available ensemble archives from European projects on similar grid size is increasing, e.g.,
CORDEX (Evans, 2011). The goal of dynamic downscaling, i.e., to extract local-scale
information from large-scale GCM data, is achieved by developing and using the limited area
models (LAMs) or regional climate models (RCMs) (Tang, Xu and Xu, 2012).

2.10. Climate Scenarios

Climate scenarios are plausible representations of the future that are consistent with
assumptions about future emissions of greenhouse gases and other pollutants and with our
understanding of the effect of increased atmospheric concentrations of these gases on global
climate. Scenarios have been used by decision makers and planners to analyze the situations
where outcomes are uncertain (Bjørnæs, 2015). In climate change research, they have become
an important element as they allow researchers to understand the long-term consequences (van
Vuuren et al., 2011) and describe plausible pathways of climatic condition and other aspects

18
of future (Moss et al., 2010). The goal of working with scenarios is not to predict the future
but to better understand uncertainties and alternative futures, in order to consider how robust
different decisions or options may be under a wide range of possible futures (Wayne, 2013).
Future greenhouse gas concentrations are an unknown because we cannot predict what
activities humans will engage in that will reduce or increase them. The choice of climate
scenarios and related non-climatic scenarios is important because it can determine the outcome
of climate impact assessment (IPCC-TGICA, 2007). A range of scenarios can be used to
identify the sensitivity of an exposure unit to climate change and to help policy makers decide
on appropriate policy responses and recently, the new scenarios called Representative
Concentration Pathways (RCP) were developed and used for preparing fifth assessment report
(AR5) of IPCC released in 2015.

2.10.1. CORDEX-Africa Domain

World climate Research program has initiated Global Coordinated Regional Downscaling
Experiment with the intention of producing an ensemble of high resolution climate change
projections by downscaling GCM simulations from Coupled Model Inter-comparison Project
Phase 5 (CMIP5) data archive (Taylor et al., 2012) on 25Km and 50Km grid spacing for 14
regions of the Globe. Africa was selected as the first target region for the CORDEX activity:
see (Jones et al., 2011) for more details on CORDEX Climate Projection Frame-work.
CORDEX-Africa RCMs generate an ensemble of high resolution historical and future climate
projections at regional scale by downscaling different GCMs forced by RCPs based on the
Coupled Inter-comparison Project Phase 5 (CMIP5).

Output form CORDEX-Africa domain have been used for impact studies especially over east
Africa such as one carried out by (Ngaina, et al., 2015) they found out that all the CORDEX
models performed well in simulating rain fall over east Africa and predicted high variation of
minimum and maximum temperature for both RCP4.5 and RCP8.5. However, temperature
increase was found to be higher during long rainy season while increase in rainfall was
found to be higher during short rainy season. Nikulin et al., (2012) evaluates the ability of ten
RCMs over Africa and concludes that all RCMs simulate the seasonal mean and annual cycle
quite accurately. Likewise, it is verified that the mean of multi-model outputs do better than
individual simulation. Hernández-Díaz et al., (2013) strengthens the achievement of (Nikulin

19
et al., 2012). They successfully reproduce the overall features of geographical and seasonal
distribution over most Africa. A further study carried out in the southern Africa on extreme
precipitation events and future climate, found out that the simulated climate output correlates
well with the observation of station data (Pinto et al., 2016).

2.10.2. Regional Climate Model (RCM)

Regional climate models (RCMs) have become important tools for providing detailed
estimates of future climate change, to support a range of studies concerned with the regional
impacts of climate change on natural ecosystems, society and even government and large
utility planning (Abbasa et al., 2016). The spatial resolution of Regional Climate Models
(RCMs) provides more reliable results on the regional scale as compared to General
Circulation model (GCMs) which provide a new opportunity for climate change effects
analysis (Nikulin et al., 2018). The regional climate model requires the driving force of the
global climate model to downscale the global climate data to the local area. Many studies of
regional effects of different global warming levels are based on ensembles of coarse-resolution
global models, however, for impact assessment and adaptation planning, high-resolution
regional climate information is necessary (Osima et al., 2018).

Africa is one of the most vulnerable continents to weather and climate variability ((IPCC,
2007). Due to its vulnerability to climate variability, the significant impacts of projected
climate change, and the general lack of climate projections based on Regional Climate
Downscaling tools, Africa was selected as the first target region for the CORDEX activity: see
Jones et al., (2011) for more details on CORDEX Climate Projection Frame-work. The
CORDEX downscaled Regional Climate Model (RCM) has relatively fine scale or resolutions
(50km x 50km). (Teutschbein and Seibert, 2012) provided a recent review on the use of RCMs
for hydrological models. They recommend that a bias correction is necessary for using the
outputs in any hydrological models as RCMs are susceptible to systematic model errors
caused by imperfect conceptualization, discretization and spatial averaging within grid cells.
These biases are typically due to the occurrence of too many wet days with low-intensity rain
or incorrect estimation of extreme temperature in RCM simulations

20
3.2.1.1. Comparisons of SRES and RCP climate scenarios

Reducing of income alterations between world regions technology is the main driving force
for demographic and economic development is assumed in all SRES scenarios. In four
scenarios change in technological, demographic and economic are driving force of the future
the greenhouse gas and Sulphur emission. The four RCP scenarios used in CMIP5 lead to
radiative forcing values that range from 2.6 to 8.5 W m –2 at 2100, a wider range than that of
the three SRES scenarios used in CMIP3 (IPCC, 2013). The SRES scenarios do not assume
any policy to control climate change, unlike the RCP scenarios. RCP4.5 and SRES B1 have
similar radiative forcing at 2100, and comparable time evolution (within 0.2W m –2). The
radiative forcing of SRES A2 is lower than RCP8.5 throughout the 21 st century, mainly due to
a faster decline in the radiative effect of aerosols in RCP8.5 than SRES A2, but they converge
to within 0.1 W m–2 at 2100 (IPCC,2013). RCPs represent pathways of radiative forcing, not
linked with exclusive socio-economic assumption in contrary to Special Report on Emission
Scenarios (SRES). Any single radiative forcing pathway can result from a diverse range of
socio-economic and technological development scenarios (Van Vuuren et.al, 2011). Two key
differences between the new RCPs and the previous scenarios (1) there are no fixed sets of
assumptions related to population growth, economic development, or technology associated
with any RCP. (2) The RCPs are spatially explicit and provide information a global grid at a
resolution of approximately 60 kilometers. This gives the spatial and temporal information
about the location of various emissions and land use changes. This is an important
improvement as the location of some emissions affects their warming potential

3.2.1.2. Representative Concentration Pathways (RCP)

Representative concentrations pathways (RCPs) have been defined as a basis for long term
and near term climate modeling experiments in the climate modeling community (Van
Vuurent et al., 2011). The definitions of the RCPs allows for a parallel development of new
socioeconomic, technical, and policy scenarios that provide insights into the impact of policy
decisions on the future climate (Van Vuurent et al., 2011). RCPs are time and space dependent
trajectories of concentrations and emission of greenhouse gases and pollutants resulting from
human activities, including changes in land use. Different types of emission scenarios are used

21
in climate studies to assess the long-term impact of atmospheric greenhouse gases and
pollutants based on assumptions of population growth, economic development level, etc. A
new generation of scenarios was developed and later used within CMIP5 of the
intergovernmental Panel on climate change (IPCC) Fifth assessment Report (RR5) published
on 2013-2014. The latest scenarios developed by the research community are denoted by
Representative Concentration Pathways (RCPs); (Van Vuuren et al., 2011). The Fifth
Assessment Report of IPCC, the scientific community has defined a set of four new scenarios,
denoted RCPs. The RCPs represent the four range of GHG emissions scenario include a
severe mitigation scenario (RCP2.6), two intermediate scenarios (RCP4.5 and RCP6.0), and
one scenario with very high GHG emissions (RCP8.5). Each RCPs defines a specific
emissions trajectory and subsequent radiative forcing (Wayne, 2013)

RCP2.6 (Low emissions) is developed by the IMAGE modeling team of the PBL Netherlands
Environmental Assessment Agency (Wayne, 2013). Radiative forcing level first reaches a
value of around 3.1 W/m2 by mid-century, and returns to 2.6 W/m 2 by the end of 21st centuries
(van Vuuren, Stehfest, et al., 2011). In order to reach such radiative forcing levels, greenhouse
gas emissions are reduced substantially, over time (Characteristics quoted from van Vuuren et
al., 2011). The important assumption in this scenario is that cropping area increases faster than
current trends, while grassland area remains constant. Forest vegetation continues to decline at
current trends.

RCP4.5 (Intermediate emissions) developed by the GCAM modeling team at the Pacific
Northwest National Laboratory‘s Joint Global Change Research Institute (JGCRI) in the
United States. It is the stabilization scenario (Wayne, 2013) in which radiative forcing
stabilizes at 4.5W/m2 (approximately 650 ppm CO2-equivalent) in 2100 without ever
exceeding that value (Thomson et al., 2011). The RCP4.5 stabilization scenario is a cost-
minimizing pathway. The major assumptions of this scenario are the global population reaches
a maximum of 9 billion by 2065 and then declines to 8.7 billion in 2100, declines in energy
consumption, increase in fossil fuel consumption, substantial increase in renewable energy and
nuclear energy use, decreasing use of croplands and grasslands due to yield increases and
large increase in forest area due to strong reforestation programs as a mitigation strategy (Wise
et al.,. 2009).

22
RCP6.0 (Intermediate emissions) developed by the AIM modeling team at the National
Institute for Environmental Studies (NIES) in Japan. It is stabilization scenario were radiative
forcing stabilizes at 6.0 W/m2 in the year 2100 without exceeding that value in prior years
(Masui et al., 2011). In this scenario, the GHG emissions will be the highest in 2060 and then
decline thereafter. The primary assumptions of this RCP are increase in energy demand, shift
from coal based to gas based production technologies, increase in use of non-fossil fuel energy
type and increase in population and economic growth in urban area, expansion of cropland and
forest area, and decrease in grassland (Masui et al., 2011).

RCP8.5 (High Emission Scenarios) consistent with a future with no policy changes to reduce
emissions (Bjørnæs, 2015). It was developed by Integrated Assessment Framework by the
International Institute for Applied System Analysis (IIASA) in Australia using MESSAGE
model (Wayne, 2013) and characterized by increasing greenhouse gas emissions that lead to
high greenhouse gas concentrations over time (Riahi et al., 2011). This scenario is highly
energy intensive with total consumption continuing to grow throughout the century reaching
well over three times current levels. Oil use grows rapidly until 2070 after which it drops even
quickly. Land use continues current trends with crop and grass areas increasing and forest area
decreasing which is driven by an increase in population, a world population of 12 billion by
2100, lower rate of technology development, heavy reliance on fossil fuel, high energy
intensity and no implementation of climate policies (Riahi et al., 2011).

2.11. Hydrological Model

Hydrological modeling is a powerful technique of hydrological system investigation for both
the research hydrologist and practicing water resources engineers involved in the planning and
development of integrated approach for the management of water resources. Hydrological
modeling methods are widely used to quantify the effects of climate and land use land cover
change (Tekalegn et al., 2018). A hydrological model involves the application of
mathematical equations to model the physical response of a watershed to meteorological
events in a watershed area (Rwigi, 2014). Hydrological model uses mathematical estimations
of stream flow as a function of basin characteristics with rainfall and drainage area being the
most important input element in addition to soil characteristics, land cover, topography of the

23
watershed, soil water content and availability of an aquifer therefore hydrological models are
vital and necessary for water and environment management (Abdulkareem et. al., 2018).

They have been developed for many different reasons and therefore have many different
forms. However, hydrological models are in general designed to meet one of the two primary
objectives. Among the objectives of hydrological models, getting a better understanding of the
hydrologic processes and how changes occur in a watershed, predicting hydrological processes
and providing important information for studying potential impacts of changes like LULC are
the major ones (Kassa, 2007).

2.11.1. Type of Hydrological Model

A. Lumped models: parameters in these hydrologic models typically do not vary spatially
within the basin and thus, basin response is evaluated only at the outlet, without explicitly
accounting for the response of individual sub-basins. This model empirically based on
integrating over some time and space scales. The high non-linear response of watersheds
appears linear. Parameters of lumped models often do not represent physical features of
hydrologic processes and usually involve a certain degree of empiricism. Though they are
conceptually simple and easy to program and apply, they require substantial calibration
data and they are not useful outside of the range of calibration. Examples include
WATBAL, SRM, HYRROM, SWM4, IHACRES, and others.
B. Distributed models: Parameters of distributed models are fully allowed to vary in space at
a resolution usually chosen by the user. These are broken down into small time and space
increments. The physical processes occurring in the watershed may be explicitly simulated
and then integrated to produce the watershed response. Examples in this category are
MIKE-SHE, Wetspa, CASC2D, WATFLOOD, and others. They can be used to analyze
changing conditions such as land use changes, project alternatives, and climate change and
they are extendible beyond the calibration range. However, they are data intensive and
code development is difficult. Distributed models generally require large amounts of (often
unavailable) data for parameterization in each grid cell. However, the governing physical
processes are modeled in detail, and if properly applied, they can provide the highest
degree of accuracy.

24
C. Semi-distributed models: parameters of semi-distributed models are partially allowed to
vary in space by dividing the basin into a number of smaller sub-basins. They are a
mixture of empirical and physically -based approaches. They lie between lumped and
distributed models. They have the advantage of increased spatial resolution and better
process descriptions over simple lumped parameter models and have computational
advantage over fully distributed, physics-based models. Some of the examples a Semi-
distributed model includes SWAT, HEC-HMS, HPV, TOPMODEL, and others.

2.11.2. Application of SWAT Model

The applicability of Soil and Water Assessment Tool (SWAT) model in the impact of land use
land cover dynamics on stream flow and sediment yield from Upper Awash basin from Koka
Dam was used for analyzing sediment yield. Therefore, 45.05 m3/s of stream flow and 57.06
Mtone annual sediment yield were entered to Koka dam during 1990 and 66.56 m 3/s of stream
flow and 47.36 Mtone annual sediment yield was extracted from the upper awash basin during
2013 land use land cover data. In addition, urban area was increased by 166%, while
plantation area appeared to be highly increased. The result obtained shows that there is 47.75
% increment of stream flow and 17 % decrement of sediment yield in 2013 as compared to
1990 land use/land cover data by (Tarik, 2017).

Dagenent et al (2017) Analysis of the combined and single effects of LULC and climate
change on the streamflow of the Upper Blue Nile River Basin (UBNRB): Using statistical
trend tests, remote sensing land cover maps and the SWAT model. The results show an
increase of evapotranspiration by up to 0.84%, 59.8% and 55.5% under LULC, climate and
combined climate and LULC change by the end of the 21st century under RCP8.5 compared
to the baseline period, respectively. Furthermore, both stream-flow and lateral flow are
projected to increase by up to 12.85% (9.9%), 28.5% (20.03%) and 26.4% (29.12%) under
LULC, climate and combined climate and LULC change scenarios, respectively. As predicted,
the shift in magnitude in RCP8.5 emissions is greater than RCP2.6 and RCP4.5.

Birhan et al., (2021) Applied SDSM and SWAT model for modeling projected impacts of
climate and land use/land cover changes on hydrological responses in the Lake Tana Basin,
upper Blue Nile River Basin, Ethiopia. As predicted, the shift in magnitude in RCP8.5

25
emissions is greater than RCP2.6 and RCP4.5. The impacts of climate change on water
balances are relatively higher than the combined effects of changes in climate and LULC.
Future LULC shifts, on the other hand, change comparatively offsetting hydrological
components.

Aich et al., (2014) this study aims to compare impacts of climate change on stream flow in
four large representatives African river basins: the Niger, the Upper Blue Nile, the Ouban-gui
and the Limpopo. Eco-hydrological model SWIM (Soil and Water Integrated Model)

Accordingly, the study concluded Kinati et al., (2019) SWAT model can be used investigate
the Effect of Land Use Land Cover and Climate Change on River Flow and Soil Loss in
Didessa River Basin, South West Blue Nile, Ethiopia Future climate changes under RCP
scenarios enhanced river discharge and soil loss in the river basin. Average monthly river flow
increased by 4.9, 5.7 and 10.6 m3/s due to LULCC between1986 and 2001, 2001 and 2015,
and in the long-term between 1986 and 2015, respectively.

26
 3. MATERIALS AND METHODS

3.1. Description of Study Area

3.1.1. Location

The Yadot river watershed is located in Bale Zone, Oromia Regional State South-eastern part
of Ethiopia. The watershed is located in Genale-Dawa basin at the upper most parts of the
Deyu sub basin. Geographically, the Yadot watershed is located between 6° 18’N and 6°51’N
latitudes and 39°49’E and 39°58’E longitudes in South Eastern part of Ethiopia shown on
(Figure 3.1). Yadot River originates from an elevation of 4,373 meters above mean sea level,
in the Bale Mountains and drains to an elevation of 946 meters above mean sea level at the
outlet of the watershed, and total area of the watershed is 735.6 km 2.Yadot River is gauged
and its minimum and maximum flows are known and the river is perennial but the flow is
decreased during winter season and increased during summer season.

27
 Figure 3.1: Location map of Yadot Watershed
3.1.2. Climate

Ethiopian climate is mainly controlled by the seasonal migration of the Intercontinental

Convergence Zone (ITZ) which is conditioned by the convergence of trade winds of the
northern and southern hemisphere and associated atmospheric circulation. It is also influenced
by regionally and locally by the complex topography of the county. The inter-tropical
convergence zone (ITCZ) is the major rain causing mechanism in Ethiopia. The movement of
ITCZ in the northward direction brings moisture from the South Atlantic Ocean, which results
in the high rainfall in the study area. The traditional climate classification of the country is
based on altitude and temperature Wurch (cold climate at more than 3000 m altitude), Dega
(temperate like climate-highland with 2500-3000 m altitude), Woina-Dega (warm 1500-2500
m altitude), Kola (hot and arid type, less than 1500 m in altitude), and Bereha (hot and hyper-
arid type) climate (NMSA, 2001). According to this classification, the study area falls in
Wurch, Dega, Woina Dega and Kola climate zone. Distributions of basin precipitation and
temperature are strongly related to the altitude with increasing precipitation and temperature
decreasing with altitude. Altitude, topography and vegetation influence climate greatly,
resulting in micro climate in specific localities and macro climate in larger areas.

A. Rainfall

The southern regions of Ethiopia experience two distinct wet seasons, which occur as the
ITCZ passes through this to its southern position. The entire Genale river sub-basin falls under
the “bi-modal” rainfall regime with two wet seasons. The major rainfall occurs from March to
May with peak flow in May and again there is rainfall from September to November with peak
rain fall in October. The mean annual rainfall (1985-2015) of the study area as shown in
(Figure 3.2) varies from around 975 mm Rira up to 1059 mm for Delo Mena and mean annual
areal rainfall of the watershed is 1029 mm (Figure 3.3).

28
 Average Monthly Rainfall (mm)
250

200

150

100

50

0
JAN Feb MAR APR MAY JUN JUL AUG SEP OCT NOV DEC

Time period (Month)

Delo mena Rira

Monthly Areal Rainfall (mm)

Figure 3.2: Mean Monthly Rainfall Distributions of the Stations (1985-2015)

250

200

150

100

50

0
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC

Time Series (Month)

Areal Rainfall

Figure 3.3: Mean areal rainfalls of the watershed from (1985-2015)

A. Temperature

It is widely agreed that in the region distribution of temperature is strongly reliant on

elevation. Previous studies on Genale Dawa River Basin show that the predicted rise of
temperature with a decreasing elevation was closer to 0.64ºC per 100 m elevation (Tsegazeab,
2014). Latitude is a secondary factor influencing mean monthly temperature. The Mean annual
maximum temperature at the study varying from 21.6 0C at Rira to 28.720C in Delo Mena and
mean annual minimum temperature in the sub basin varies from 9.78 0C to 15.620C at Rira and

29
Delo mean, respectively. The mean annual temperature varies from 12.7 0C to 25.160C in the
study period (Figure 3.4).

35
Temerature (oC)

30
25
20
15
10
5
0
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
Time Series (Month)
TMP Max Delo Mena TMP MAX Rira Mean TMP Delo Mena
TMP MIN Delo Mena TMP MIN Rira Mean TMP Rira
F
igure 3.4: Average Monthly Max, Min and Mean Temp in the period of (1985-2015)
3.1.3. Land Use Land Cover
The main land use land covers of the watershed are agricultural lands, grass/range land, forest
land, scrub/bush land, wood land and settlement. Out of the total watershed areas 735.6 km2
(about 60.45%) is covered with dominant forest land and 14.21 % cover with agriculture based
on land use land cover map 2015 (Table 4.4) and Figure (4.1).

3.1.4. Soil
Soil data is a significant component for land use land cover change impact on hydrological
components of watershed management. According to Food and Agricultural Organization –
Harmonized World Soil Database (FAO-HWSD) soil classification, the soil mapping units
have been identified in the study area at a scale of 1:1250, 000. The soil type identified in the
study area encompasses eight soil types’ and the dominant soil type were pellic vertisols and
chromic vertisols covers about 37.84% and 26.42% of the watershed total area, respectively
(Appendices Table 2).

30
Figure 3. 5: Soil map of the study watershed

3.2. Data Collection Methods and Procedures

3.2.1. Data Collection

3.2.1.1. Digital Elevation Model

Digital Elevation Model (DEM) of 12.5m x 12.5m resolution was processed for the extraction
of flow direction, flow accumulation, stream network generation and delineation of the
watershed and sub-basins. The topographic parameters of the study watershed, such as terrain
slope, channel slope or reach length was also derived from the DEM.

31
 Fig
ure 3.6: Digital Elevation Model of Yadot watershed
3.2.1.2. Land Use Land Cover Data

Land use land cover is one of the main input data of the SWAT model to analyze the impacts
on stream flow of the watershed and describe the Hydrological Response Units (HRUs) of the
watersheds. The historical LULC images were obtained from Landsat images and classified
using supervised classification in Earth Resource Data Analysis System (ERDAS) imagine
model. Therefore, the classified 1985, 2000 and 2015 and the predicted future LULC 2035 and
2055 maps was based on the classified historical satellite images using IDRISI Selva v.17
model in the study watershed. The SWAT model has predefined four letter codes for each land
use category. These codes were used to link or associate the land use map of the study area to
SWAT land use databases. Hence, while preparing the lookup-table, the land use types were
made compatible with the input needs of the model.

32
3.2.1.3. Soil Map Data

Soil properties are one of the major inputs data required by SWAT2012 model of the
watershed. The soil map of the study area obtained from Food and Agricultural Organization –
Harmonized World Soil Database (FAO-HWSD) and cross checked with HWSD Viewer soil
map. According to this soil database there are 8 soil types in the study watershed. To integrate
the soil map with SWAT model, a user soil database which contains textural and chemical
properties of soils was prepared for each soil layers and added to the SWAT user soil
databases using the data management append tool in ArcSWAT10.4.1.

3.2.1.4. Hydro – Meteorological Data

Weather data required for SWAT input includes daily data of precipitation, maximum and
minimum temperature, relative humidity, wind speed and solar radiation. For this study only
two station within and around Yadot watershed are Delo Mena and Rira station considered due
to total absence recorded data in other station and absence of the influence remaining station
in Thiessen Polygons. The climate data used for this study cover 31 years from January 1985
to December 2015. Rainfall and temperature data are available in all the stations whereas
sunshine hours, wind speed and relative humidity data were not available for Rira stations.
The Delo Mena station was used as weather generator station so as to generate weather
variables for missing records assigned with no data code value (-99) in SWAT database. The
climate data for study periods was finally prepared in text file format and imported to the
SWAT model database.

The stream flow data of the Yadot watershed is needed for the calibration and validation of the
model. The daily stream flow data (1985-2008) at near Delo Mena station collected from the
Minister of Energy and Water Resources of Ethiopia for the Yadot watershed. The missing
flow data were filled by linear regression from the nearby river gauged station.

33
3.2.1.5. Checking Hydro-Meteorological Data quality

A. Filling of missing Precipitation data

Hydro-meteorological data missing is a common challenge to perform hydrological analysis

and modeling to use long time series data and filling the missing data is very crucial. Failure
of any rain gauge or absence of observer from a station causes a short break in the record of
rainfall at a given station. Methods used to fill missing data such as: arithmetic mean method,
normal ratio method and inverse distance weighing method. Arithmetic mean method can be
used to fill in missing data when normal annual precipitation is within 10% of the
gauge/station for which data are being reconstructed. The normal ratio method is used when
the normal annual precipitation at any of the index station differs from that of the precipitation
station by more than 10%. The Arithmetic mean method is given as (equation number 8).

Px= ∑ n Pi
i=1 -------------------------------------------------------------------------------------------8
n ¿
¿

Where: - Px is the missing precipitation value for station i (i=1, 2, 3...n), and P1, P2… Pn are
precipitation values at the adjacent stations for the same period and n is the number of nearby
stations. The normal ratio method is as given (equation number 9).

1
n
Pi
Px= ∑ ¿ Nx−−−−−−−−−−−−−−−−−−−−−−−−−−−−−¿ ¿9
n i=1 ¿

where PX is the missing precipitation value for station X for a certain period of time, Pi is
precipitation values at adjacent stations for the same period, NX is the long-term, annual
average precipitation values at station X; Ni is the long-term precipitation for neighboring
stations and n is the number of adjacent stations.

B. Checking the Consistency of Data

Checking for inconsistency of the record is done by the double-mass curve technique
(Subramanya, 2013) for this study. The principle of double mass analysis is to plot cumulative
values of the station under investigation against cumulative values of another station, or
cumulated values of the average of other stations, over the same period of time. Double mass

34
curve is a graphical method for identifying and adjusting inconsistency in a station record by
comparing its time trend with those of adjacent stations. The problem occurs when the
watershed rainfall data at the rain gauge station is inconsistent over a period and adjustment of
the measured data is necessary to provide a consistent record. If significant change in the
regime of the curve is observed, it should be corrected by using (equation number 10).

Px∗Mc
P cx= ------------------------------------------------------------------------------------------10
Ma

Where: Pcx is corrected precipitation at any time period, Px is original recorded precipitation
at time period, Mc is corrected slope of the double mass curve and Ma is original of the slope
of double mass curve.
- Se
Cumulative Annual Precipitation of

35000

30000
lected Station in (mm)

25000

20000

15000
Rira
10000 Linear ( Rira)
Delo Mena
5000

0
0 5000 10000 15000 20000 25000 30000 35000

Cumulative Annual Precipitation of Group Station in (mm)

Figure 3.7: Double Mass Curves for the Stations

C. Climate Data Homogeneity Test

The total annual rainfall data of all stations those selected in and around the sub basin was
homogenous as verification by the study analysis (Figure 3.8) and (Table 3.4).

35
 Delo Mena Rira
1500 1300

1300
1100
Total Annual Rainfall (mm)

Total Annual Rainfall (mm)

1100
900
900
700
700

500 500
1980 1985 1990 1995 2000 2005 2010 2015 2020 1980 1985 1990 1995 2000 2005 2010 2015 2020
Year Year

Mean Annual Rainfall Mean Annual Rainfall

F
igure 3.8: Homogeneity tests of annual rainfall data of individual rain gauge stations
Table 3.1: Summary of Homogeneity test climate data with 5% (α = 0.05) significance level

No Station Minimum Maximum Mean Std. deviation

1 Delo Mena 822.54 1404.18 1059.42 163.11
2 Rira 730.10 1268.66 975.30 147.53
The annual stream flow (m3/year) of Yadot River near Delo Mena gauging station was
checked for homogeneity test for the period of 1985 – 2008 as shown in (Figure 3.9) below.

5000
Yadot Annual Flow(m3/s)

4500
4000
3500
3000
2500
2000
1500
1000
500
1985 1990 1995 2000 2005 2010
year

Yadot Annual Flow mu = 2581

Figure 3.9: Pettitt's Homogeneity test of Annual Stream flow

36
3.2.2. Land Use Land Cover Change Detection

3.2.2.1. Image Processing and Date of Acquisition

In the study of the impacts of land use land cover and climate change on stream flow of the
watershed, remote sensing images are required and can be processed by computers to produce
land use land cover map. Data from remote sensing sensors sometime came with sort of errors.
Errors may not be used for mapping without correction. Different data pre-processing
techniques were used such as; image enhancement, geometric and radiometric correction were
implement to prepare land use map. Land sat 5 TM, Landsat7 ETM+ and Land Sat8 OLI/TIRS
were selected for the period of 1985, 2000 and 2015 respectively (Table 3.2). Each land sat
was geo-referenced to WGS_84 datum and Universal Traverse Mercator (UTM) Zone 37N.
To avoid a seasonal variation in vegetation pattern and distribution throughout the study year,
the selection of dates of the acquired data were made as much as possible in the same annual
season of the acquired years. In order to view and discriminate the surface features clearly, all
the input satellite images were composed using the true (RGB) colour composition and false
colour composition to identify images provide complete coverage of Yadot sub basin

Table 3.2: Land sat images used for LULC change analyses and their characteristics

Sensor Path/Row Date of acquisition Resolution Season

Land sat 5 TM 167/055 14/01/1985 30x30m Dry
167/056 14/01/1985 30x30m Dry
Land sat 7 ETM+ 167/055 18/01/2000 30x30m Dry
167/056 18/01/2000 30x30m Dry
Land sat 8 OLI 167/055 03/03/2015 30x30m Dry
167/056 03/03/2015 30x30m Dry
3.2.2.2. Image Classification

Image classification is a process of assigning pixels to predefined land use land cover classes
or done in order to assign different spectral signatures from the LANDSAT datasets to
different LULC. The main objective of image classification is to place all pixels in an image
into LULC classes in order to draw out useful thematic information.

37
 In remote sensing, there are various image classification methods. There appropriateness
depends on the purpose of land use land cover maps produced for and the analyst’s knowledge
of using the algorithms. However, in most cases the researchers categorized them into three
major categories: supervised, unsupervised and hybrid. For this research, supervised
classification approach was adopted so as to have acceptable classification accuracy. In this
system the user should have a good knowledge about the land cover to be studied. In
supervised image classification system, the user relies on her/his prior knowledge and skills
and can select a group of pixels belongs to a particular land use land cover.

The supervised classification was applied after defined area of interest (AOI) which is called
training sample. The training site were selected in agreement with the Land sat Image and
connects with Google Earth, then the samples for each land cover type were merge together.
Using Google Earth historical Image, topographic map and original mosaic image as a
reference compared with the corresponding classification. In supervised classification,
defining of training sites, extraction of signature editor and classification of image was
performed using Maximum Likelihood classifier. Accordingly, six categories of LULC were
identified based on the nature and similarity of the use and cover types. This classification is
in agreement with (Desalegn et al. 2014; and Adane and Getachew, (2017) in Bale eco-region.

Agriculture: - This land use encompasses areas that allocated for crop production of both
annuals and perennials,

Grass/Range land: - Either communal or private grazing lands that are used for livestock
grazing. The land is basically covered by small grasses, grass like plants and herbaceous
species. It also includes area covered with range or grass used for grazing, as well as little
grassland with some bare lands or includes other small sized plant species

Forest: - Land area covered with a high density of trees or dense trees and nearly closed
canopies which include evergreen forest and plantation forests.

Scrub/Bush land: - Land area covered by Asta scrubland, Erica bushes, alpine vegetation and
small white leaves found at top of Sanette Platue and habitats of Ethiopian Wolf.

38
Wood Land: - Land not classified as forest, spanning more than 0.5 hectares include
deciduous forest and mixed forest land.

Settlement: - Land feature where there is a permanent populated urban settlement, densely
concentrated rural settlement and other facilities.

3.2.2.3. Accuracy Assessment

Accuracy assessment is the comparison of a produced classification with what actually exists
on mapped image for the same location to evaluate how well the classification results. It is
performed by comparing a map created by using remote sensing analysis to a reference map
based on different information sources such as Google Earth and original mosaic images.

The classification accuracy assessment is based on the samples collected from the image, the
generating of enhanced images and the computation of the probability in which this pixel
belong to this class. Total test samples of 150, 180 and 228 for the image 1985, 2000 and 2015
were randomly selected from the original mosaic images for the validation of 1985 and 2000
and Google earth images for 2015 with ERDAS IMAGINE 2015, respectively. Evaluation of
the accuracy of a classified image can be done using an error matrix sometimes called
confusion matrix. An error matrix is a square array of rows and columns and presents the
relationship between the classes in the classified and reference data. The columns normally
represent the reference data, while the rows indicate the classification generated from
classified image. Error of omission refers to pixels in the reference map that were identified as
something other than their "accepted" value. Whereas error of commission, on the other hand
refers to pixels that were incorrectly classified as a class in a row.

Overall accuracy, user’s accuracy and producer’s accuracy were derived to test the
classification. The overall accuracy provides accuracy of the whole image classification
(number of correctly classified pixels divided by the total number of pixels in the error
matrix). Most of the classification accuracy measurements are derived from an error matrix.
Based on this, user’s accuracy refers to the probability that a given pixel can be found in the
ground as it is in the classified image, whereas producer’s accuracy refers to the percentage of
a given class that is correctly identified on the map (Sahalu, 2014).

39
3.2.3. Land Use Land Cover Change Detection Analysis
Land use land cover change detection is the process to determine a change in the classified
class of satellite images by looking at different periods of studies. Change detection entails
findings the type, amount and location of land use changes. In this study five time periods
LULC change detection has been made i.e. for 1985, 2000 and 2015 for historical and 2035
and 2055 for future period. These time periods were chosen based on the resettlement
program, the availability of quality satellite image and other data like weather and
hydrological data in the study area. The LULC change detection was performed using post
classification cross-tabulation approach in ArcGIS software. The classified images were
compared in the three time periods and change statistics was computed by comparing the
values of area of one LULC data set with the corresponding value of the next data set for each
period (using equation 4). The annual rate of change of each land use land cover was
calculated using the following formula (number 5).

The comparison of the LULC statistics assisted in identifying the percentage change, trend and
annual rate of change between 1985 and 2015. In achieving this, the first task was to develop a
table showing the area in hectares and the percentage change for each past LULC 1985, 2000
and 2015 and for future LULC 2035 and 2055 measured against each LULC type.

Final Year Area – Initial Year Area

Percentage Change ¿ ……………………….…………….
Initial Year Area
….4

FinalYear Area – Initial Year Area

Annual Rate of Change= ……………………………….…
Number of Year Between two image
5

3.2.4. Land Use Land Cover Prediction and Model Validation

3.2.4.1. Land Use Land Cover prediction

Land Use Land Cover change predictions for 2035 and 2055 periods were employed using the
CA-Markov model, which is available in IDRISI; Geospatial software for monitoring and
modeling the Earth system; version 17.0. The Cellular Automata (CA) and Markov Chain
models are considered to be advantageous for modeling land use changes (Mishra and Rai,

40
2016; Parsa et al., 2016). CA-Markov, when compared with other dynamic models has
advantages such as the ability of individuals to boost information that has under the regular
principles, regulates their evaluation, and analysis of the spatial (space) characteristics, used to
easily predict land use in the future (Gambo et al., 2018). CA-Markov predicts both the trend
and the spatial structure of different LULC categories (Wang et al., 2012; Li et al., 2015)
based on the basis observed LULC image, transition probability matrix and suitability images
as a group file (Clark labs, 2012; Eastman, 2012). The minimum level for accuracy assessment
in identification of LULC categories in remote sensing data should be at least 85%.
Afterwards, the data is exported into an ASCII text file to enable for further analysis in
ArcGIS version 10.4.1 .This study employed the 2015 classified map as a basis LULC image,
and the 2000 and 2015 maps for assembly transition probability matrix (with the proportion
errors of 0.15). The land use maps of 2000 and 2015 were taken as a baseline map and CA-
Markov in IDRISI software was used to predict future land use maps of LULC 2035 and
LULC 2055.

3.2.4.2. Validating LULC Prediction Model

The CA-Markov chain is a stochastic procedures model that pronounces the possibility of
change from one land use land cover class into another land use land cover class using a
transition probability matrix. In this study, IDRISI Selva v.17 is used to predict the future
LULC of study area on CA-Markov model. CA-Markov will use the 1985 and 2000 maps to
produce a simulated 2015 map, which is important to validate with actual LULC of 2015 map
through KIA (Kappa Agreement of Index) approach (Mishra and Rai, 2016; Parsa et al.,
2016). Any model prediction requires model calibration and validation. Then, the simulated
LULC map of 2015 was compared with the observed 2015 LULC map using the Kappa index:

P o−P c
K appa= ----------------------------------------------------------------------------------------6
1−P c

In the above equation, Po is the proportion of correctly simulated cells; Pc is the expected
proportion correction by chance between the observed and simulated map. If Kappa ≤ 0.5
shows rare agreement, 0.5 ≤ Kappa ≤ 0.75 shows a medium level of agreement, 0.75 ≤ Kappa

41
≤ 1 shows a high level of agreement and Kappa = 1 for perfect agreement (Pontius and
Malanson, 2005).

3.2.5. Near-term and Mid-term Climate Projection

3.2.5.1. GCMs and RCMs used for this study

In this study, the dynamically downscaled RCM product of daily precipitation, maximum and
minimum temperatures outputs from the fifth phase of the Coupled Model Inter comparison
Project (CMIP5) over the CORDEX-Africa domain at spatial grid resolution of 0.44 o (~50
Km) were used. The CORDEX-Africa program provided climate data that were simulated by
three GCMs (HadGM2-ES, MPIESM-LR and ICHEC-EC) and dynamically downscaled by
three RCMs, which are Regional Climate Limited-area modeling (CCLM), Rossby Center
regional atmospheric model (RCA4) and KNMI Regional Atmospheric Climate Model,
version 22 (RAMO22T), respectively for each GCM (Table 3.3). The selected RCMs run on
SWAT with two RCP scenarios of the high emission scenario of RCP8.5 and mid emission of
RCP4.5. All Regional Climate Models have simulated seasonal mean annual cycles of
precipitation and temperature with a significant bias shown on individual models. However,
the use of ensemble mean over the individual RCMs helps to minimize the highest and lowest
projections and enables to minimize the uncertainties by working with the average of the
RCMs. The reference to this study was (Wakjira et al., 2019) in their study of evaluation of
the CORDEX regional climate models performance in simulating climate conditions of two
watersheds in Upper Blue Nile Basin.

For this study, climate model selection was based on the previous study on Africa by other
researchers and literature (Lennard et al., 2018: Samuel et al., 2019) and Hussen et al., 2013)
over East Africa. Hussen et al., (2013) conducted Assessment of the Performance of 10
CORDEX Regional Climate Models in Simulating East African Rainfall and reported; apart
from REMO and CCLM most of the model showed bias under wet and dry season. Samuel et
al., (2019) on comparison of 30 GCM models outputs reported that CanESM2 and HadGEM2-
ES performed best in terms of the correlation coefficient between gauged and simulated
rainfall amounts. Therefore, for the climate scenarios, the daily inputs of precipitation and
temperature (maximum and minimum) were obtained from downscaled projections for two
42
representative concentration pathways of RCP4.5 and RCP8.5. The future scenario was made
up to 2080 years for near and midterm in this study and the impact was analyzed per 30 years.
The RCMs climate data obtained from CORDEX-RCM is in a NetCDF format with a rotated
pole of (longitude and latitude) then extracted on ArcGIS10.4.1. The unit of precipitation and
temperature were kg/m2/s and Kelvin, respectively and there is a need for correcting the
rotated pole and units before using the data in any impact analysis. Therefore, to obtain
precipitation in mm/day the given data was multiplied by 86400 (24 * 60* 60) and to obtain
temperature in Celsius by subtract 273.15 from downscaled data using the above conversion
factors.

Table 3.3: Summaries of GCM and RCM climate model used in the study

Full Name GCM Full Name of RCM Climate Center

HadGEM2-ES GCM Model

Resolution
Spatial
Model

Hadley Global Regional Climate 0.440 Met Office Hadley

Environment Limited-Area Center
Model 2-Earth Modeling
CCLM

System

Coupled Model Rossby Center 0.440 Max Planck Institute

Version 5, Medium Regional for Meteorology
MPI-ESM-LR

Resolution Atmospheric Model (MPI-M)

RCA4

Irish Center for KNMI Regional 0.440 EC-EARTH

RACMO22T

High-End Atmospheric Consortium

EC-EARTH

Computing Earth Climate Model,

Consortium version 2.2

3.2.5.2. Bias Correction of Future Climate Data

Bias correction is widely used in climate impact aims to adjust selected statistics of a climate
model simulation to better match between observed meteorological data. Statistical bias
corrections often need to be performed to better match the model output to the observations

43
(piani et al., 2010). Bias correction is usually needed as climate models often provide biased
representations of observed time series data due to systematic model errors caused by
imperfect conceptualization, discretization and spatial averaging within grid cells.

Power transformation: - uses the exponential form pb that is used to adjust the statistical
variance standard deviation and coefficient of variation as well as the mean of precipitation
time series. In this nonlinear correction each daily precipitation amount P is transformed to a
corrected P* using (equation number 7).

b
P∗¿ a P −−−−−−−−−−−−−−−−−−−−−−−−−−−−−−7

Where: P* is the bias corrected daily precipitation; P is the uncorrected daily precipitation; and
a and b are the transformation coefficients. The determination of the b parameter is done
iteratively so that, for areal of each grid box in each month, the coefficient of variation of the
corrected RCM daily precipitation time series matches that of the observed precipitation time
series. Then a is the coefficient that is determined from the mean of observed precipitation
data and the mean of Pb. Finally, monthly constants a and b are applied to each uncorrected
daily observations corresponding to that month in order to generate the corrected daily time
series

Variance Scaling: - to correct temperature time series, as temperature is known to be

approximately normally distributed (Yang et al, 2015). The VARI method was developed to
correct both the mean and variance of normally distributed variables such as temperature
(Teutschbein and Seibert, 2012). Temperature is normally corrected using the VARI method
(equation number 8).

σ (¿ )
Tc= [ T−µ ( T ) ] x + µ ( ¿ ) −−−−−−−−−−−−−−−−−−−−−−−−¿8
σ (T )

Where
Tc: the corrected daily temperature; T: the uncorrected daily temperature from RCM model;
σ(To) is the standard deviation of the observed temperature; σ(T) is the standard deviation of
the uncorrected temperature; µ(T) is the simulated mean temperature and µ(To) is the
observed mean temperature.

44
3.2.6. Impact of Climate and LULC Change on Stream flow
3.2.6.1. Climate Change Impact Simulation Using SWAT Model

The projected precipitation, temperature, baseline and future land use land cover maps were
used as input to the calibrated and validated SWAT model by keeping constant wind speed,
solar radiation and relative humidity throughout the future simulation period to determine
stream flow in the future time period. The baseline period consists of the years 1985–2015 and
future simulation include two future period (2021-2050) and (2051 -2080) for near and
midterm. Three simulation scenarios were developed for the independent and combined
effects of the Land Use Land Cover Change and climate change. The first scenario (S1)
considers Only Land Use Land Cover Change effects on stream flow using land use maps of
1985, 2000, 2015, 2035 and 2055 with baseline climate data from 1985 – 2015 for stream flow
simulation. The second scenarios (S2) considers only a climate change effects on stream flow
using single baseline land use map of 2015 with three future climate scenario data of 1985–
2015, 2021–2050 and 2051–2080 for stream flow simulation. The third scenario (S3)
considers combined effect of climate and Land Use Land Cover change on stream flow using
land use maps of 2015 as baseline period, LULC map of 2035 for near-term, while LULC map
of 2055 for the midterm simulation with baseline and future climate data of 1985–2015, 2021–
2050 and 2051–2080, respectively.

3.2.6.2. Description of SWAT Model

Hydrological models have been developed for different reasons and different forms. It is
mathematical descriptions of the different components of the hydrologic cycle. However,
hydrological models are generally developed to meet the primary objectives of enabling better
understanding of the hydrologic processes and for hydrologic prediction in a watershed.

The reasons, for selecting SWAT model for this study are:-

It has been applied in studies related to land use land cover and climate changes in
different parts of the world including Ethiopia and proved for sufficiency of performance.
The model can simulates the required hydrological process in the study watershed
Less demanding of the input data is Advantageous.

45
 It is readily and freely available.

The Soil and Water Assessment Tool (SWAT) is an open-source-code, semi-distributed model
with a large and growing number of model applications in a variety of studies ranging from
watershed to continental scales (Arnold et al., 2012, Neitsch et al., 2011). It evaluates the
impact of LULC and climate change on water resources in a basin with varying soil, land use
and management practices over a period of time (Arnold et al., 2012). SWAT model has the
ability to simulate and predict the long-term changes of various hydrological elements under
different land use types, soil types and management practices on the large scale complex
basin. SWAT hydrological models has been successfully used in the studies of the impact of
land use land cover and climate change on water resources in a basin with varying soil, land
use and management practices over a set period of time (Arnold et al., 2012). It is evident that
the SWAT model has yielded high accuracy for short/long-term simulations of yearly and
monthly mean streamflow (Zuo et al., 2016; Anand et al., 2018).

In SWAT, the watershed is divided into multiple sub-basins, which are further subdivided into
hydrological response units (HRUs) consisting of homogeneous land use management, slope
and soil characteristics (Arnold et al., 1998: Arnold et al., 2012). HRUs are the smallest units
of the watershed in which relevant hydrologic components such as evapo-transpiration,
surface run-off and peak rate of run-off, groundwater flow and sediment yield can be
estimated. Water balance is the driving force behind all the processes in the SWAT calculated
using (equation 9).

t
SW t =SW 0+ Σ n=1( R ¿ ¿ day−Q surf −E a−W seep−Qgw ) … … … … … … … … … … … ..9 ¿

Whereas, SWt is the final soil water content in [mmH 2O], SW0 is the initial soil water content
on day i [mm], t is the time [days], R day is amount precipitation on specific days i [mmH 2O],
Qsurf is the amount of surface runoff on specific day i [mmH 2O], Ea is amount evapo-
transpiration on a day i [mmH2O], Wseep is the amount of water entering the vadose zone from
soil profile on day i [mmH2O] and Qgw is amount of return flow on a day i [mmH2O].

46
3.2.6.3. Data Input for SWAT Model

For this study, Digital Elevation Model (DEM) of 12.5m by 12.5m was downloaded from
https://vertex.daac.asf.alaska.edu/ to delineate the watershed of the study area. DEM was used
in the SWAT model along with soil and land use land cover data to delineate the watershed
and to further divide the watershed into sub watersheds and hydrologic response units (HRUs).
The soil map of the study area was obtained from Ministry of Water, Irrigation and Energy of
Ethiopia. Remote sensing satellite data used for this study were obtained from the United
States Geological Survey (USGS) Earth Resources Observation and Science (EROS) data
center (https://earthexplorer.usgs.gov/) website in GEOTIFF file format. Satellite data include
Land sat 5 TM, Land sat 7 ETM+, and Land sat 8 OLI images for the year 1985, 2000 and
2015. Earth Resource Data Analysis System (ERDAS) imagine version 15 and ArcGIS10.4.1
software packages were utilized for the classification of LULC changes.

The climate model data were obtained from outputs of high resolution regional climate models
of CORDEX RCM dynamically downscaled by the regional climate model are (CCLM4-8-17,
RACMO22T and RCA4) https://climate4impact.eu/impactportal/data/esgfsearch.jsp.

Daily climate variables include daily precipitation [mm], maximum and minimum temperature
[ºC], solar radiation [MJ/m2/day], wind speed [m/s] and relative humidity [-]) were obtained
from National Meteorological Service Agency of Ethiopia (NMSA), while Observed Daily
stream flow data of Yadot River at Delo Mena gauging Station were obtained from Ministry
of Water, Irrigation and Electricity (MoWIE) .

3.2.7. SWAT Model Setup

3.2.7.1. Watershed Delineation

The watershed and sub watershed delineation was performed using 12.5m resolution DEM
data using Arc SWAT model watershed delineator tools. First, the SWAT project set up was
creating necessary folders and databases to store all the data. The watershed delineation
process consists of five major steps, DEM setup, stream definition, outlet and inlet definition,
watershed outlets selection and definition and calculation of sub basin parameters. Once, the
DEM setup was completed and the location of outlet was specified on the DEM, the model

47
automatically calculates the flow direction and flow accumulation. Finally stream networks,
sub watersheds and topographic parameters were calculated using the respective tools.

The stream definition and the size of sub basins were determined by selecting the threshold
area or minimum drainage area required to form the origin of the streams to be two thousands
(2000 hectares) in order to decrease sub basin number. The Yadot watershed was delineated
into 23 sub basins having an estimated total area of 735.6km2.

3.2.7.2. Hydrological Response Unit (HRU)

After watershed delineation, land use, soil and slope characterization for watershed was
performed using commands from the HRU analysis menu on the Arc SWAT Toolbar. The
HRU analysis tool in Arc SWAT helps to load land use, soil layers and slope map to the
project. The HRUs represent areas with homogeneous land use/cover, management, and soil
characteristics. After HRUs created promptly two common options were provided in the
model to define the HRUs. The first was by assigning a single HRU and the other was with
multiple HRUs option.

In this study, therefore, the multiple HRUs option was employed aiming at obtaining reliable
results of simulation. The SWAT user’s manual suggests that a 20 % land use threshold, 10 %
soil threshold and 20 % slope threshold are adequate for most modeling application. However,
(Setegn et al, 2008), suggested that HRU definition with multiple options that account for 10%
land use, 20% soil and 10% slope threshold combination gives a better estimation of runoff
and sediment components. Therefore, for this study, HRU definition with multiple options that
accounts for 10% land use, 20% soil and 10% slope threshold combination was used to
eliminate minor land use and land covers in sub basin, minor soil within a land use and land
cover area and minor slope classes within a soil on specific land use and land cover area. Due
to this the Yadot Watershed was divided into 185 HRUs, each has a unique land use and soil
combinations.

3.2.7.3. Sensitivity Analysis

Sensitivity analysis is the process of determining the rate of change in model output with
respect to changes in model parameters (Arnold et al., 2012). Sensitivity analysis provides for

48
better understanding of the behavior of the system being modeled, such as model parameters
and applicability, thus it increases the confidence level of the model and its predictions.
Therefore, prior to calibration and validation process, sensitivity analysis was carried out to
reduce the number of parameters that needs optimization. In this study semi-automated
Sequential Uncertainty Fitting (SUFI 2) was used to identify the sensitive parameters. Global
analysis requires large number of simulations (Arnold et al, 2012) which can also be a
problem. However, in this research the number of simulations used for calibration was 1000,
which is large enough to get accurate results for global sensitivity analysis. The average
monthly stream flow data of 15 years from 1988 to 2002 of the watershed gauging station
were used to compute the sensitivity of the stream flow parameters.

The aim of sensitivity analysis is to estimate the rate of change in the output of a model with
respect to changes in watersheds that result in a clear difference in hydrologic sensitivity. In
the sensitivity analysis process, by using SWAT-CUP, first create the new folders and copy
the SWAT simulation results to perform the sensitivity analysis and the location of the sub
basin where observed data was compared against simulated output then, chosen parameters
were entered for the sensitivity analysis with the default lower and upper parameter bounds.

3.2.7.4. Model Calibration

Calibration is a major aspect of hydrological modeling which intended to fit the simulated
outputs of the model to the recorded stream flow data of the watershed by adjusting the model
parameters, thereby reducing the prediction uncertainty. Parameters for modification are
selected from those identified by sensitivity analysis. Additional parameters other than those
identified during sensitivity analysis are used primarily for calibration due to the hydrological
processes naturally occurring in the watershed.

Therefore combination of both manual and automatic or numerical parameter optimization

method was used in this particular study. First the parameters were calibrated manually for the
period of fifteen year (1988 – 2002) and warm-up period of three year (1985-1987) until the
model simulation results acceptable as per the model performance measures. Next, the final
parameter values that would be manually calibrated were used as the initial values for the auto
calibration procedure. The graphical and statistical approaches were also be used to evaluate

49
the SWAT model performance a number of times until the acceptable values were obtained for
surface runoff independently.

3.2.7.5. Model Validation

Validation is the comparison of the model outputs with independent data set without making
any adjustment. The purpose of model validation is to check whether the model can predict
flow for another range of period. In order to utilize any predictive watershed model for
estimating the effectiveness of future potential management practices the model must be first
calibrated to measured data and should then be tested (without further parameter adjustment)
against an independent set of measured data. The model was validated using stream flow from
2003 to 2008 at the location of study area for six year.

3.2.7.6. Model Performances Evaluation

The goodness of fit was evaluated with Nash-Sutcliffe efficiency (NSE), coefficient of
determination (R2), root mean square error standard deviation ratio (RSR), and percentage bias
(PBIAS) were used for model evaluation for quantification of accuracy in watershed
modeling.

Coefficient of Determination (R2), a measure of consistency of simulated and observed data,

was also considered in evaluating the model. The closer the value of R2 to one implies a
perfect agreement between the simulated and observed flow (Singh et al., 2005).

R 2=
∑ ( Xi−Xav )∗∑ (Yi−Yav) ---------------------------------------------------------------------
∑ √ ( Xi−Yav ) 2 ∑ √ (Yi−Yav ) 2
-----14

Where Xi = measured value (m3/s); Xav = average measured value (m 3/s); Yi = simulated
value (m3/s) and Yav is average simulated value (m3/s)

The Nash-Sutcliffe efficiency coefficient (Nash and Sutcliffe, 1970) is used to assess the
predictive power of the hydrological models. The Nash-Sutcliffe simulation efficiency
indicates how well the plot of observed and simulated value fits. When the value of NSE is 1,
the simulated value is the same as observed (perfect). When the value is between 0 and 1, it

50
indicates the deviations between observed and simulated data. An efficiency of lower than
zero indicates that the mean value of the observed time series would have been a better
predictor than the model (Krause et al., 2005). The Nash-Sutcliffe efficiency (NS) is
calculated by using equation 15:

ENS=1−
∑ ( Xi−Yi ) 2 … … … … … … … … … … … … … … … … … … … … … … … … … … … … …. .
∑ ( Xi−Xav ) 2
15

Where Xi is measured value; Yi is simulated value and Xav is average observed value.

Table 3.4: Performance ratings of recommended statistics for monthly stream flow

Performance R2 NSE PBIAS (%) RSR

rating
Very good 0.86 < R2 ≤1 0.75 < NSE ≤1 PBIAS < 10 0.0<=RSR<=0.5
Good 0.75 < R2 ≤ 0.86 0.65 < NSE ≤ 0.75 10 ≤ PBIAS < 15 0.5<RSR<=0.6
Satisfactory 0.65 < R2 ≤ 0.75 0.5 < NSE ≤ 0.65 15 ≤ PBIAS < 25 0.6<RSR<=0.7
Un satisfactory R2 ≤ 0.65 NSE ≤ 0.5 PBIAS ≥ 25 RSR>=0.7

51
 4. RESULTS AND DISCUSSION

4.1. Land Use Land Cover Classification Accuracy Assessment

The accuracy assessment is used to determine the degree certainty classified image. The
results from accuracy assessment showed an overall accuracy obtained from the random
sampling process for the image of 1985, 2000 and 2015 were 87.33%, 91.67% and 92.25%,
respectively. The overall accuracy of LULC maps of 2015 show higher than 1985 and 2000
this may be due to higher spatial resolution of satellite images and clear visibility on Google
earth. Therefore, the validation data set indicate excellent agreement of the classified image
with ground truths as shown on (Table 4.1, 4.2 and 4.3).

Table 4.1: Confusion Matrix for LULC Classification 1985

REFERENCE DATA
LULC AG GL FR SB WL ST Total UA
AG 22 1 0 1 0 0 24 91.67%
GL 2 22 0 0 1 0 25 88%
FR 0 0 35 1 3 0 39 89.75%
SB 0 4 0 20 1 0 25 80%
WL 0 0 4 1 26 0 31 83.87%
ST 0 0 0 0 0 6 6 100%
Total 24 27 39 23 31 6 150
PA 91.67% 81.5% 89.75% 86.96% 83.87% 100% K= 0.843
Overall Classification Accuracy = 87.33%
Note: Agriculture = AG, Grass/Range Land = GL, Forest = FR, Scrub/Bush Land = SB,

Wood Land = WL, Settlement = ST, Producer Accuracy = PA, Users Accuracy = UA

The error matrix for 1985 reveals that settlement was most often accurately classified in that
year of imagery followed by agriculture land, forest, wood lands, grass/range land and
scrub/bush land (Table 4.1). Some areas covered by Grass/range, scrub/bush and forest land
were sometimes mistaken for wood land in 1985. User accuracy for each classes range from
91.67% for agriculture land to 100% for settlement.

52
Tables 4.2: Confusion Matrix for LU/LC Classification of 2000

REFERENCE DATA
LULC AG GL FR SB WL ST Total UA
AG 28 2 0 0 0 1 32 90%
GL 1 19 0 0 1 0 21 90%
FR 0 0 50 0 3 0 53 94.3%
SB 0 2 1 25 0 0 28 89.3%
WL 0 0 2 1 33 0 36 91.7%
ST 1 0 0 0 0 10 11 90%
Total 30 23 53 26 37 11 180
PA 93.33% 82.6% 94.3% 96.1% 89.2% 90% K = 0.896
Overall Classification Accuracy = 91.67%
Table 4.3: Confusion Matrix for LU/LC Classification of 2015

REFERENCE DATA
LULC AG GL FR SB WL ST Total UA
AG 32 2 0 0 3 0 38 84.2%
GL 0 22 0 0 3 0 25 88%
FR 0 0 83 1 0 0 84 98.8%
SB 0 1 0 31 1 0 33 93.94%
WL 0 2 3 0 25 0 30 83.33%
ST 1 0 0 0 0 18 19 94.74%
Total 33 27 86 32 32 18 228
PA 81.48 K = 0.904
97% 96.51% 96.88% 78.13% 100%
%
Overall Classification Accuracy = 92.25%
The LULC map 2000 classification error matrix indicates that Forest was correctly classified,
followed by wood land, Agriculture, Grass/range land, settlement and scrub/bush land as
shown on (Table 4.2). In the 2000 classification, agriculture and scrub/bush land was
commonly mistaken for grass/range land whereas grass/range and forest land mistaken for
wood land. User accuracy for each classes range from 89.3% for scrub/bush land to 94.33%
for forest land.
53
The 2015 classification error matrix reveals that forest was accurately classified, followed by
settlement, scrub/bush, grass/range, agriculture land and wood land as shown (Table 4.3). In
the 2015 classification agriculture, grass/range and scrub/bush land was commonly mistaken
for wood land similarly agriculture, scrub/bush and wood land mistaken for grass/range land.
The reason why forest land accurately classified is that because the forest land are well known
and easily classified. The User’s accuracy reflects the reliability of the classification and the
more relevant measure of the classification’s actual utility in the field.

4.2. Land Use Land Cover Change Analysis

The major land use land cover types shown on the maps of 1985, 2000, and 2015 include
agriculture land, grass/range land, Scrub/Bush land, settlement, forest and wood land
presented in (Table 4.4). Forest land comprised largest share of total area about 65.50%
followed by wood land 13.29% based on 1985 LULC map which indicates that much of the

area was covered by green vegetation. The 1985 historical reference LULC map show that

7.54%, 7.45%, 6.11% and 0.12%, of grass/range land, scrub/bush, agriculture and settlement

land area, respectively. The analysis has shown that the agricultural land was increased from
6.11% to 14.21%, while settlement land increased from 0.12% to 0.73% between 1985 and
2015, respectively. The annual rate of agricultural land change showed 235.55ha/year,
161.94ha/year and 192.33 ha/year increment from 1985 to 2000, 2000 to 2015 and 1985 to
2015, respectively. Settlement revealed annual rate of increment by 15.43ha/year,
14.54ha/year and 14.50ha/year for the period 1985 to 2000, 2000 to 2015 and 1985 to 2015,
respectively. Grass/range land area cover was increased from 7.54% to 7.7% between 1985 and
2000, while declined to 7.3%, from 2000 to 2015 of the total area of watershed. In the first
scenarios the land under grass/range land increased by (2.15%), while in the second scenarios
decreased by 5.25% and third scenarios grass/range land decline by 3.21%, respectively.

54
 Table 4.4: The area coverage of LULC, percent, and rate of changes in the Ydaot watershed between 1985, 2000 and 2015

Area (ha) Percentage change (%) Rate of change (ha/year)

LULC 1985 2000 2015

1985 -2000

1985 -2000
2000-2015

1985-2015

2000-2015

1985-2015
Agriculture 4491.01 (6.11*) 8024.2(10.91*) 10453.4(14.21*) 78.67 30.27 132.76 235.55 161.94 192.33
Forest 48180.4(65.5*) 46415.1(63.1*) 45200.1(60.45*) -3.66 -2.62 -6.19 -117.68 -81.00 -96.14
Grass/Range 5544.65(7.54*) 5663.72(7.7*) 5366.55(7.3*) 2.15 -5.25 -3.21 7.94 -19.81 -5.75
Scrub/bush 5476.35(7.45*) 5401.33(7.34*) 5253.04(7.14*) -1.37 -2.75 -4.08 -5 -9.89 -7.20
Settlement 87.83(0.12*) 319.23(0.43*) 537.41(0.73*) 263.47 68.34 511.88 15.43 14.54 14.50
Wood land 9773.05(13.29*) 7729.63(10.51*) 6742.77(9.17*) -20.91 -12.77 -31.01 -136.23 -65.79 -97.75
*The number in parenthesis/bracket indicate percentage LULC

Key: - First scenarios (S1) (1985 to2000), Second scenarios (S2) (2000 to2015) and Third scenarios (S3) (1985 to 2015)

55
The annual rate of grass/range land change showed 7.94ha/year increments from 1985 to
2000, while declining by 19.81ha/year and 5.75ha/year between 2000 to 2015 and 1985 to
2015 period, respectively. Forest, woodland and scrub/bush land shows a declining trend for
entire study periods from 1985 to 2015 with the magnitude of 6.19%, 31.01% and 4.08%. In
the first scenarios the land under forest decreased by 3.66%, wood land by 20.91% and the
land under scrub/bush land declined by 1.37%, respectively. The result for the second
scenarios indicate that land under forest, wood land and scrub/bush land continued to decrease
by 2.62%, 12.77% and 2.75%, respectively. The annual rate of decline of forest, wood land
and scrub/bush land was 117.68ha/year, 136.23ha/year and 5ha/year in the first scenarios,
while for the second scenarios declined by 81ha/year, 65.79ha/year and 9.89ha/year,
respectively (Table 4.4). The classified images were obtained after pre-processing and
supervised classifications as shown below on figures 4.1.

Figure 4.1: LULC Map of Yadot Watershed in the period 1985, 2000, and 2015

56
The increase of grass/range land may due to shifting cultivation practices contributed for
conversion of other LULC to grass/rangelands. On the other hand fallowing is one way of
farmland management practices in the study area, implying that the fallow farmlands used as a
grazing land for cattle. The increasing trend of agriculture and settlement is due to the
expansion of human needs and population growth. Generally, the result indicates that forest
land, wood land, scrub/bush land and grass/range land decreased, while agricultural land and
cultivated settlement increased. This may be due to the incasing of population growth rate and
the need to secure more farm land. The findings of the study are consistent with other studies
conducted in Ethiopia such as. Similarly (Dires and Temesgen, 2020) reported that agricultural
land and settlement areas increased from 1984 to 2018 in the Lake Tana Basin. This result was
lies with (Tatek and Danial, 2019) revealed a significant increasing trend of agriculture and
settlement areas by 12 and 270%, respectively. In contrast, grasslands, forest lands and shrub-
bush lands showed a declining trend of about 40, 21 and 12%, respectively in the Muga
watershed. Similarly (Mesfin, 2018) reported that increasing settlement and agriculture in
shaya watershed between1987 and 2015. Study (woldeamlak and Solomon, 2013) reported
expansion of open grassland, agriculture and settlements in a tropical highland watershed,
Ethiopia as a result of human activities such as the reduction of natural vegetation cover.

4.3. LULC Change Transition Probability Matrix between 1985 to 2015

Conversion between 1985 and 2000

The amount of LULC in hectare converted to other or gained from other type are important
aspect of change detection. Agriculture gained about 3290.62ha, 1157.22ha, 121.26ha,
10.01ha and 2.49ha of land from wood land, grass/range land, forest, scrub/bush and
settlement land, respectively. Conversely about 515.76ha, 395.31ha, 98.03ha, and 39.31ha of
agricultural land lost to wood land, grass/range land, settlement and forest, respectively. The
change detection indicted that the shifting of (524.56ha) forest land, (566.46ha) scrub/bush
land, (1597.82ha) wood land, (2.86ha) settlement and (395.31ha) agricultural land had
significantly changed to grass/range land. Moreover, the area of wood land (82.39ha), (0.06ha)
scrub/bush land, (98.03ha) agriculture and (56.28ha) grazing land had changed slightly into a
settlement. Gains and losses in forest, scrub/bush land and wood land were also taken place during
these periods (Table 4.5).

57
Table 4.5: Transition Area Matrix (ha) between 1985-2000 and 200-2015 period in Yadot watershed.

Land Use Land Cover 2000

Land use class Agriculture Forest Grass/Range Scrub/Bush Settlement Wood Total
Agriculture 3442.61 39.31 395.31 0.00 98.03 515.76 4491.01
Forest 121.26 46103.29 524.56 230.79 0.00 1200.46 48180.37
LULC1985

Grass/Range 1157.22 23.94 2576.72 358.39 56.28 1372.10 5544.65

Scrub/Bush 10.01 3.74 566.46 4692.66 0.06 203.42 5476.35
Settlement 2.49 0.00 2.86 0.00 82.48 0.00 87.83
Wood Land 3290.62 244.84 1597.82 119.49 82.39 4437.88 9773.05
Total 8024.20 46415.13 5663.72 5401.33 319.23 7729.63 73553.25

Land Use Land Cover 2015

Agriculture Forest Grass/Range Scrub/Bush Settlement Wood Total
Agriculture 5885.81 0.54 961.88 0.00 126.57 1049.40 8024.20
Forest 148.04 45006.10 52.59 20.59 4.36 1183.45 46415.13
LULC2000

Grass/Range 1683.82 145.61 2786.33 177.65 71.98 798.33 5663.72

Scrub/Bush 75.02 0.00 176.13 5005.51 0.09 144.58 5401.33
Settlement 40.67 0.00 6.06 0.00 267.01 5.50 319.23
Wood Land 2620.00 47.86 1383.57 49.29 67.39 3561.51 7729.63
Total 10453.35 45200.12 5366.55 5253.04 537.41 6742.77 73553.24

58
Conversion between 2000 and 2015

The conversions of LU/LC from one class to another class were revealed in all study periods . Between
2000 and 2015 period, 2620 ha, 1683.82 ha, 148.04 ha, 75.02ha and 40.67 ha of wood lands,
grass/range lands, forest, scrub/bush land and settlement were converted to agriculture land,
respectively. Similarly, 1049.40ha, 961.88ha, 126.57ha and 0.54ha agriculture land were lost
to other LULC categories. In these periods, a significant area of wood land was converted
from forest (1183.45 ha), agriculture 1049.40ha, settlement (5.50), scrub/bush land (144.58ha)
and grass/range land (798.33 ha). In reverse, there was also a considerable conversion of wood
land to other categories. During this period, some areas of settlement were also converted from
farmland (126.57 ha), forestland (4.36 ha), wood land (67.39ha), scrub/bush land (0.09ha) and
grassland (71.98 ha). Although, about 5.50 ha, 40.67 ha, and 6.06 ha of the settlement were
also converted to wood land, agriculture and grass/range land, respectively. Gains and losses
in forest, scrub/bush land and grass/range land were taken place in this study periods. A
significant amount of gains and losses in the grass/range land has occurred in these periods.
The gain of LULC for each class was determined from the result of persistence and the total
column value, whereas the loss is from total row and the persistence (figure 4.2).

Loss Gain Persistence Loss Gain Persistence

50000
60000
45000 35000
30000
15000 20000

0
5000
-15000 Agriculture Forest Grass/ scrub/Bush Settlement Wood land
Range land
Land -10000 Agriculture Forest Grass/ scrub/Bush Settlement Wood land
Range Land land

Area (Ha)
Area(Ha)
Figure 4.2: Gain, Loss and Persistence area of LULC class in (1985 -2000) and (2000-2015)
The gains and losses of the land use and land cover during the study period were derived from
the cross tabulation of 1985, 2000 and 2015.

59
4.4. Future Land Use Land Cover Change prediction

4.4.1. Validation of CA-Markov Model

In order to validate the LULC prediction given by the CA-Markov model, the simulated land
use areas were used to compare the actual and simulated land use areas. Then, the observed
LULC 2015 was cross-compared with the simulated LULC 2015 to evaluate the performance
of the model using the kappa index. The result on (Table 4.6) indicates agricultural,
grass/range land and settlement have best agreement (Figure 4.3).

Table 4.6: Comparison of actual and projected LULC 2015

Simulated Classified
LULC Types Area (ha) Percent (%) Area (ha) Percent (%)
10324.8 14.04
Agriculture 10453.4 14.21
5175.2 7.04
Grass/Range Land 5366.55 7.7
43,760.61 59.49
Forest 45200.1 60.45
5902.1 8.02
Scrub/Bush Land 5253.04 7.14
7865 10.69
Wood Land 6742.77 9.17
526.3 0.72
Settlement 537.41 0.73
Total 73,553.31 100 73,553.31 100

The simulated LULC map shows that forest land coverage areas are underestimated, while
wood land and Scrub/Bush Land are overestimated. The values of k-index greater than 80%
show good agreement between the projected and actual LULC map.

Table 4.7: The k-index values of the simulated LULC map of 2015

Index Value
K no 0.891
K locations 0.9016
K location Strata 0.9016
K standard 0.84

60
All indices are greater than 80%, showing a good overall agreement and projection ability of

the model. Therefore, CA-Markov model is strong to simulate the future for accurate
prediction of future LULCs

Figure 4.3: Simulated and observed LULC Map of 2015

4.4.2. Predicted future Land Use Land Cover
The future prediction of LULC 2035 and 2055 were executed using the 2015 land use map as
a base map. The predicted LULC result of 2035 reflects that agricultural, settlement,
grass/range land and wood land are increasing, while forest and scrub/bush land are
decreasing. The LULC between the two future periods 2035 and 2055 indicate agricultural
land and settlement predicted identical increasing trend, while forest, scrub/bush land, wood
land and grass/range land showed a decreasing trend as shown on (Figure 4.4).

The predicted LULC result of 2035 and 2055 indicate agricultural and settlement showed
identical increasing trend from 2015 to 2055 by 44.02% and 69.20% respectively. The annual
rate of agricultural and settlement increment from 2015 to 2035 showed 89.34 ha/year and
11.66 ha/year, while from 2035 to 2055 increased by 140.74 ha/year and 6.94ha/year,

61
respectively. Generally agricultural and settlement lands observed through the period 2015 to
2055 increased by 115.04 ha/year and 9.30ha/year, respectively.

Figure 4.4: Predicted LULC Map of the year 2035 and 2055
Based on the LULC prediction result grass/range lands were increased by 32.38% between
2015 and 2035 at the expense of other LU/LC categories mainly woodland, scrub/bush land
and agricultural lands as shown on (Table 4.8). Shifting cultivation practices contributed for
conversion of woodlands to rangelands in the lowland parts of the watershed creating huge
openings in woodland after cultivation is abandoned or due to continuous increase of future
conservation works. The grass/range land cover slightly declined by 1.53% from 2035 to 2055
but greatly increased by 30.35% from 2015 to 2055. The LULC result in this study agrees with
other previous studies in Ethiopia. For example (Bireda, 2015) reported that expansion of
grassland between 1973 and 2015, almost 80% of was obtained from cultivated lands in Fagita
Lekoma Woreda, Awi Zone, North Western Ethiopia.

62
 Table 4.8: The area coverage of LULC, percent, and rate of changes in the Ydaot watershed between 2015, 2035 and 2055

Area (ha) Percentage change (%) Rate of change (ha/year)

LULC 2015 2035 2055

2015 -2035

2015 -2035
2035-2055

2015-2055

2035-2055

2015-2055
Agriculture 10453.4(14.21*) 12240.1(16.64*) 15054.9(20.47*) 17.09 23 44.02 89.34 140.74 115.04
Forest 45200.1(60.45*) 36945.1(50.23*) 35467.6(45.22*) -18.26 -4 -21.53 -412.75 -73.88 -243.31
Grass/Range 5366.55(7.3*) 7104.26(9.66*) 6995.56(9.51*) 32.38 -1.53 30.35 86.89 -5.44 40.73
Scrub/bush 5253.04(7.14*) 4690.07(6.38*) 4671.12)6.35) -10.72 -0.4 -11.08 -28.15 -0.95 -14.55
Settlement 537.41(0.73*) 770.55(1.05*) 909.27(1.24*) 43.38 18 69.2 11.66 6.94 9.30
Wood land 6742.77(9.17*) 11803.1(16.05*) 10454.9(14.21*) 75.05 -11.42 55.05 253.02 -67.41 92.80
*The number in parenthesis/bracket indicate percentage LULC

Key: - First scenarios (S1) (2015 to 2035), Second scenarios (S2) (2035 to 2055) and Third scenarios (S3) (2015 to 2055)

63
Similarly (Afera, 2018) also reported expansion of grassland at the expense of built up area
and water body of Woreta Zuria, Ethiopia Watershed between 1997 and 2017.

The forest and scrub/bush land cover showed continuously declining trend from 2015 to 2055
by 21.53% and 11.08% with annual rate of decline by 243.31ha/year and 14.55 ha/year,
respectively. Wood land showed increasing trend from 2015 to 2035 by 75.05% but slightly
declined by 11.42% from 2035 to 2055. The annual rate of wood land increment showed
253.02 ha/year from 2015 to 2035, while declined with annual rate of 67.41ha/year for the
period 2035 to 2055, respectively. In general increment of wood land from 2015 to 2035 may
be due to the planned afforestation and green legacy strategies of government are the possible
reason. (Semegnew et al., 2021) summaries forestland was declined to 77.8% in 2017,
whereas farmland, settlement, and grassland increased by 17.4%, 3.4%, and 1.4%,
respectively from 2002 to 2017. Similarly (Adane and Getachew, 2017) reported that overall
increasing grass/range land from2006 to 2026 in Bale eco-region. Therefore Prediction of
LULC using time serious data is important for the future management plan of LULC. Hence, a
rational land use plan should be proposed in order to sustain livelihoods of local communities.

4.5. LULC Transition Probability Matrix between 2015 to 2055

Conversions between 2015 and 2035

In these periods, 1001.34 ha, 1086.84ha, 552.32 ha, ha, 50.84 ha and 0.23 ha of agricultural
land will be converted from forest land, wood land, grass/range land, settlement and
scrub/bush land, respectively. About 159.19 ha, 66.60 ha, 64.08 ha and 0.11 ha of settlement
will be also reverted from agriculture, wood land, grass/range land and forest land,
respectively. Although agricultural and settlement land will be converted from other land uses,
there will be also a significant loss of agricultural and settlement land to other land uses (Table
4.9). Between these periods, there will be a huge loss of forest, wood land, grass/range land,
scrub/bush land to other land uses. For example about, 6499.04ha, 1001.34 ha, 755.21ha and
0.11ha of forest land will be converted to wood land, agriculture, grass/range land and
settlement, respectively. About 1086.84ha, 757.59ha, 64.08ha, 1.54ha and 0.21ha of wood
land will be also converted to agriculture, grass/range land, settlement, scrub/bush land and
forest land, respectively. The grass/range land converted to agricultural land, wood land,

64
settlement, scrub/bush land and forest land will be 552.32 ha, 173.16 ha, 66.60 ha, 6.49ha and
0.49 ha, respectively.

In contrast, wood land grass/range land, forest land and scrub/bush land will be also gain from
6970.64 ha, 2536.77ha, 8.02ha and 0.71ha other categories (Table 4.8). In the watershed, the
downstream and partly at the upstream areas, which were previously covered by forests,
scrub/bush land and range lands, have been converted to agricultural land and settlement.

Conversion between 2035 and 2055

The significant increase of agricultural land and settlement with the sharp decline of forest
land, wood land and scrub/bush land in the watershed were the major transformations
observed. For example, 1476.73ha, 760.80ha, 218.09ha, 55.44ha and 1390.60ha of agriculture
land will be reverted from forest, grass/range land, scrub/range land, settlement and wood
land, respectively. Conversely about 338.33ha grass/range land, 594.57ha wood land, 90.52ha
settlement, 63.44ha forest and 0.02ha scrub/bush land were lost from agricultural land,
respectively.

The conversions of wood land (64.90ha), agriculture land (90.52ha), scrub/bush land (0.09ha)
and grass/range land (43.76ha) was the major contributing land use for the settlement.
Conversely about 55.44ha agriculture and 5.11ha grass/range land were lost from settlement,
respectively. An estimated 1390.60ha, 1508.95 ha, 226.27ha, 57.99ha and 64.90 ha of wood
land were also converted to agriculture, forest, grass/range land, scrub/bush land and
settlement, respectively. The forest land which is the highest class has 33849.98ha with the
probability of remaining as forest land in 2035–2055. In reverse, there was also a considerable
gain and los of scrub/bush land to other categories. The major reason for the expansion of
grassland was low productivity of agriculture lands and migration of people in search food for
their cattle in the forest and wood land (Table 4.9). In 2035–2055 the two highest class loss
was the change of wood land to forest land by 1508.95ha and forest land to agriculture by
1476.73ha. The minimum loss of LULC category was observed from settlement to agriculture
55.44ha and grass/range land 5.11ha.

65
 Table 4.9: Transition Area Matrix (ha) between 2015 - 2035 and 2035 - 2055 period in Yadot Watershed

Land Use Land Cover 2035

LULC class Agriculture Forest Grass/Range Scrub/Bush Settlement Wood Total
Agriculture 9548.52 0.00 561.53 0.00 159.19 184.12 10453.35
Land Use Land

Forest 1001.34 36944.43 755.21 0.00 0.11 6499.04 45200.13

Cover 2015

Grass/Range Land 552.32 0.49 4567.49 6.49 66.60 173.16 5366.55

Scrub/Bush Land 0.23 0.00 457.97 4682.04 0.00 112.79 5253.04
Settlement 50.84 0.00 4.48 0.00 480.57 1.52 537.41
Wood Land 1086.84 0.21 757.59 1.54 64.08 4832.50 6742.77
Total 12240.10 36945.14 7104.26 4690.07 770.55 11803.1 73553.25
Land Use Land Cover 2055
Agriculture Forest Grass/Range Scrub/Bush Settlement Wood Total
Agriculture 11153.21 63.44 338.33 0.02 90.52 594.57 12240.10
Land Use Land Cove

Forest 1476.73 33849.98 583.58 0.00 0.00 1034.85 36945.14

Grass/Range Land 760.80 45.18 5814.91 170.58 43.76 269.02 7104.26
2035

Scrub/Bush Land 218.09 0.00 27.34 4442.53 0.09 2.02 4690.07

Settlement 55.44 0.00 5.11 0.00 710.00 0.00 770.55
Wood Land 1390.60 1508.95 226.27 57.99 64.90 8554.43 11803.14
Total 15054.87 35467.55 6995.56 4671.12 909.27 10454.9 73553.26
Note: The shade numbers indicate the unchanged LULC proportions

66
Generally, the patterns of LULC change showed that the agricultural land gained the most area
compared to the other LULC types.

Loss Gain Persistence Loss Gain Persistence

35000 35000

20000 20000

5000
5000
Agriculture Forest Grass/ scrub/Bush Settlement Wood land
-10000 Range Land land Agriculture Forest Grass/ scrub/Bush Settlement Wood land
-10000 Range land
Land

Area (Ha) Area (Ha)

Figure 4.5: Gain, Loss and Persistence area of LULC class in (2015 -2035) and (2035-2055)
The significant increase of agricultural land and settlement, with the sharp decline of forest
land, scrub/bush land, wood land and grass/range land experienced reduction in coverage
throughout the study periods, the greatest reduction rate was observed in forest land. In
general the study show that agriculture and settlement area increase was consistent with
(Megersa et al., 2021) reported the future projection of agricultural land increased from
61.19% in 2019 to 72.98% in 2035 and 73.24% in 2050, whereas forest cover will further
degrade from 16.94% in 2019 to 8.07% in 2050. Similarly the predicted result of the year
2019 to 2050 showed an increase of urban land from 1.2% in 2019 to 3.8% in 2050. (Birhan et
al., 2021) prediction of the LULC change using the CA-Markov chain model indicates that
cropland, tree cover, and built-up areas are likely to increase by 2020s, 2050s, and 2080s in the
Lake Tana Basin, upper Blue Nile River Basin. (Wakjira et al., 2020) reported that agricultural
land and urban and built-up continuously increase while forest land decrease throughout the
study period 1987 to 2055 respectively. Therefore, to mitigate the rapid rates of LULC
conversions at watershed, the application of integrated watershed management strategies,
managing the rapid population growth, afforestation of degraded or deforested areas, and
reducing the dependency of locals on forest products is critically important.

67
4.6. Future Climate Projection

4.6.1. Precipitation Projection under RCP4.5 and RCP 8.5 scenarios

A. Projected Rainfall under RCP4.5

The seasonal rainfall distribution under RCP4.5 and RCP8.5 scenarios were similar to the
baseline period. However, the mean annual rainfalls under RCP4.5 and RCP8.5 scenarios
slightly increased than the baseline period except in the near-term under RCP 8.5 of the
watershed. The mean annual rainfall is projected to increase by 1.38 mm and 13.26 mm for the
near and midterm period compared to baseline under RCP4.5 scenarios, respectively. The
projections of mean annual rainfall showed a slight increasing trend for near and midterm
according to scenario RCP4.5. For Scenario RCP8.5 the projected mean annual rainfall is
higher compared to scenario RCP4.5. This increase is mainly due to the projected increases in
wet season rainfall, partially offset by projected decreases in dry season rainfall (Figure 4.6).
Mean monthly Rainfall (mm)

250

200

150

100

50

0
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC

Time Series (Month)

Observed Rainfall RCP4.5 Near Rainfall RCP4.5 Mid Rainfall

Figure 4.6: Mean monthly precipitation under RCP4.5 for near and mid future compared to
baseline period.
The seasonal classifications were made based on the rainfall patterns of the study area into,
Dry season (January, February, June, July and December) and Wet season (March, April,
May, August, September, October and November), therefore, it is applicable only in south
eastern (bimodal type of rainfall) of Ethiopia. Seasonal rainfall in wet season showed
maximum increase of 12.28 mm in midterm under RCP 4.5 scenarios.

68
The projected mean monthly rainfall showed inconsistent directions and magnitude of change
under RCP4.5 and RCP8.5 scenarios. The mean monthly rainfall may increase during dry
season (January, February and July) with maximum increase in January (22.2%), while
decreased (June and December) with maximum decrease in June (15.43%) in the near-term.
Similarly in the midterm mean monthly rainfall increased in dry season (January and July)
with maximum increase in January (23.28%), while decrease (February, June and December)
with maximum decrease in June (15.87%), respectively under RCP4.5 scenarios from baseline
period. The mean monthly rainfall in the study watershed is projected to increase in the wet
season (March, September, October and November) with maximum increase in September
(16.67%), while decrease (April, May and August) with maximum decrease in August
(8.38%) in the near-term under RCP4.5 scenarios. The midterm mean monthly rainfall
increased during wet season (August, September, October and November) with maximum
increase in October (10.74%), while decreased (March April and May) with maximum
% Change Rainfall

decrease in April (12.64%), respectively under RCP 4.5 scenarios as shown on (Figure 4.7).

25
20
15
10
5
0
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
-5
-10
-15
-20
Time Series (Month)

RCP4.5 Near Rainfall RCP4.5 Mid Rainfall

Figure 4.7: Percentage changes of monthly rainfall for near-term and midterm under RCP 4.5
scenarios from baseline period
B. Projected Precipitation Under RCP8.5

The projected change in precipitation varies with mean annual and seasonal under RCP8.5
scenarios. The mean annual rainfall projection show slight decreasing trend by 4.85 mm in the
near-term, while increased by 17.56 mm in the midterm under RCP8.5 scenarios from the
baseline period. Similar to temperature, rainfall simulations under RCP8.5 is substantially
greater than under RCP4.5 in the watershed even if rainfall decrease in the near-term.

69
 Average Monthly Rainfall(mm)
250

200

150

100

50

0
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC

Time Series (Month)

Observed Rainfall RCP8.5 Near Rainfall RCP8.5 Mid Rainfall

Figure 4.8: Mean monthly precipitation under RCP8.5 for near and mid future compared to
baseline period
Under RCP8.5 the mean seasonal rainfall is projected to increase in dry and wet season. For
the RCP8.5 of the midterm rainfall expected to increase with maximum of 14.60 mm in the
wet season, while decreased by 5.69 mm in near-term of dry season from the baseline (Figure
4.8). The mean monthly change in rainfall is not systematic i.e. rainfall increases in some
months and decreases in other months. The seasonal rainfall is projected to increase during dry
season in January, February and July with maximum increase in January by 18.79% and
23.55% for near and midterm under RCP8.5 scenarios respectively. The projected rainfall
expected to decrease in June and December with maximum decrease in June by 13.52% and
14.62%, respectively for both near-term and midterm under RCP8.5 scenarios from baseline.
Similarly mean monthly rainfall increased in wet season for the month of (March, September,
October and November) with maximum increase in September by 14.03%, while decreased
for the month of (April, May and August) with maximum decrease in August by 15.14% in
the near-term under RCP8.5 scenarios from baseline, respectively.

70
 % Change Rainfall
25
20
15
10
5
0
-5 JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC

-10
-15
-20
Time Series (Month)

RCP8.5 Near Rainfall RCP8.5 Mid Rainfall

Figure 4.9: Percentage changes of monthly rainfall for near-term and midterm under RCP 8.5
scenarios from baseline period
For the midterm mean monthly rainfall expected to increase in wet season (September,
August, October and November) with maximum increase in October by 19.28%, while
decrease in March, April and May with maximum decrease in April by 5.57% under RCP8.5
scenarios from the baseline period respectively. In general future projection of precipitation
show increasing trend annually from the baseline period under RCP4.5 and RCP8.5 scenarios.
On Figure 4.9 above upward sign indicates an increase from the baseline period, while a
downward signs indicates decrease from the baseline period. There are different studies that
shows increase rainfall in the future in agreement with these results include: (Tesfaye et al,
2019) study in Awata River watershed, Genale Dawa basin indicates that precipitation does
not show systematic increase or decrease. But overall mean annual precipitation shows an
increasing trend for both RCP4.5 and RCP8.5 scenario with the percent of increment up to
26.8% and 35.1% at near (2020) and mid-term (2050) respectively. The study by (Megersa et
al., 2021) also confirmed an increasing trend in annual precipitation by +26.3 and +31.85%
for RCP4.5 and RCP8.5 climate scenarios, respectively, at the end of twenty-first century in
the Upper Wabe Bridge Watershed in Wabe Shebele River Basin. (Mesgana et el., 2017)
conclude that the mean annual precipitation is projected to increase by about 6% (9%) in
2050s (2080s) in Awash, Baro, Genale and Tekeze rivers of four major river basins in
Ethiopian. (Birhan et al, 2021) precipitation would increase by up to 25% in the Lake Tana
Basin, of the upper Blue Nile River Basin

4.6.2. Temperature Projection under RCP4.5 and RCP 8.5 Scenarios

A. Projected Temperature Under RCP4.5

71
The dynamically downscaled mean annual maximum and minimum temperature showed
increasing trend in the near and midterm under RCP4.5 and RCP8.5 scenarios. Compared to
the baseline period, seasonal (dry and wet) changes of maximum and minimum temperatures
are predicted to increase in the watershed under RCP4.5 and RCP8.5 scenarios. Therefore
baseline and projected minimum and maximum temperatures under both scenarios flows a
similar seasonal trend. The mean annual maximum temperature is projected to increase from
0.650C to 1.050C under RCP4.5 scenarios in the near-term and midterm, respectively.
Similarly the mean annual minimum temperature is expected to increase from 0.48 0C to
0.850C in the near and midterm from baseline period under RCP4.5 scenarios.
Average Monthly Tmax & Tmin (oC)

35
30
25
20
15
10
5
0
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC

Time series (Monthly)

observed Tmax RCP4.5 Near Tmax. RCP4.5 Mid Tmax.

observed Tmin RCP4.5 Near Tmin RCP4.5 Mid Tmin

Figure 4.10: Mean monthly maximum and minimum temperatures of near and mid future
period with respect to baseline period under RCP 4.5 scenarios
The simulation result showed a consistent increasing trend of mean annual for both minimum
and maximum temperature in the near and midterm with a higher rate of increase towards the
end of the century. It also noted that the rate of change of maximum temperature is higher than
the rate of change of minimum temperature. As shown in (Figure 4.10) above. The mean wet
season increase in maximum temperature shows a variation with a range from 0.650C to
1.310C, whereas the mean dry season increase in maximum temperature shows a variation
with a range from 0.650C to 0.680C in the near-term and midterm under RCP4.5 scenarios,
respectively. However, mean wet season increase in minimum temperature shows a slight
variation with a range from 0.060C to 0.650C, whereas the mean dry season increase in
minimum shows a large variation with a range from 1.070C to 1.130C in the near-term and

72
midterm under RCP4.5 scenarios respectively. The simulated maximum temperature is higher
in wet season of the midterm, while highest increase in minimum temperature is projected to
occur in dry season of the midterm in the watershed. Even though the mean annual maximum
and minimum temperature has show increasing trends in all future under RCP4.5 scenarios.
The mean monthly simulated maximum and minimum temperature shows an inconsistent
Change in Tmax & Tmin

directions and magnitude change in all months from the near to midterm of future periods
under RCP4.5 scenarios as shown below on (Figure 4.11).
6

al
Y
R

V
T

C
R
B
N

P
L

et
ry

u
A

C
A

E
U

O
P

JU
E
JA

JU

n
W
D
A

S
M

D
M
F

n
A

N
-2

A
-4
Time series (Month)

RCP4.5 Near Tmax. RCP4.5 Mid Tmax. RCP4.5 Near Tmin RCP4.5 Mid Tmin

Figure 4.11: Mean monthly and seasonal changes in maximum and minimum temperature
(°C) for near and midterm from baseline period under RCP 4.5 scenarios.
The mean monthly maximum temperature increased under RCP4.5 from 0.25 0C to 3.430C with
the maximum increase in November, while decreased from 0.55 0C to 2.320C with the
maximum decrease in March in the near-term compared to baseline period, respectively.
Minimum temperature increased under RCP4.5 from 0.49 0C to 3.640C with the maximum
increase in January, while decrease from 0.13 0C to 2.780C with the maximum decrease in May
in the near-term from baseline period, respectively. The mean monthly maximum temperature
may increase under RCP4.5 scenarios range from 1.39 0C to 5.610C with the maximum
increase in July, while decrease from 0.20C to 3.100C with the maximum decrease in February
midterm compared to baseline period respectively. Minimum temperature expected to increase
under RCP4.5 from 0.350C to 2.810C with the maximum increase in June, while decrease from
0.420C to 0.780C with the maximum decrease in December in the midterm compared to
baseline period respectively.

B. Projected Temperature Under RCP8.5

73
Mean annual maximum temperature is projected to increase from 0.71 0C to 20C, while
minimum temperature is expected to increase from 0.56 0C to 1.80C under RCP8.5 scenarios in
the near and midterm, respectively from baseline period. Compared to the baseline period, the
mean seasonal maximum temperature in dry season increased from 0.61 0C to 1.90C, while the
minimum temperature increased from 1.14 0C to 2.010C under RCP8.5 for near and midterm,
respectively as shown on (Figure 4.12).
Mean monthly Tmax & Tmin (oC)

35
30
25
20
15
10
5
0
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
Time series (Monthly)

observed Tmax RCP8.5 Near Tmax. RCP8.5 Mid Tmax.

observed Tmin RCP8.5 Near Tmin RCP8.5 Mid Tmin

Figure 4.12: Mean monthly maximum and minimum Temperature of near-term and midterm
with respect to baseline period under RCP8.5 scenarios
Similarly the projected mean seasonal maximum temperature show an increasing trend in wet
seasons from 0.780C to 2.070C, while the minimum temperature increased from 0.140C to
1.660C under RCP8.5 in the near-term and midterm, respectively relative to the baseline
period. The increase of maximum temperature is higher in wet season of midterm compared to
dry season, while the increasing of minimum temperature is larger in dry season of midterm
under RCP 8.5 scenarios.

The projected mean monthly maximum temperature has larger magnitude of increment under
RCP8.5 scenarios range from 0.090C to 2.690C with the maximum increase in the month of
July, while the maximum temperature decreasing from 0.67 0C to 1.360C with the maximum

74
decrease in February for the near-term. On the other hand, the minimum temperature increased
under RCP8.5 ranges from 0.720C to 3.740C with the maximum increase in January, while
decrease from 0.140C to 2.770C with the maximum decrease in May for the near-term,
respectively from baseline period. The mean monthly maximum temperature may increase
under RCP8.5 from 00C to 3.790C with the maximum increase in July for the midterm
Change in Tmax & Tmin

compared to baseline period, respectively as shown on (Figure 4.13).

al
Y
R

V
T

C
R
B
N

P
L

et
ry

u
A

C
A

E
U

O
P

JU
E
JA

JU

n
W
-1

D
A

S
M

n
M
F

A
-2

-3
Time series (Month)

RCP8.5 Near Tmax. RCP8.5 Mid Tmax. RCP8.5 Near Tmin RCP8.5 Mid Tmin

Figure 4.13: mean monthly and seasonal changes in maximum and minimum temperature (°C)
for near and midterm from baseline period under RCP 8.5 scenarios.
Minimum temperature may increase under RCP8.5 from 0.180C to 3.890C with the maximum
increase in June, while minimum temperature decrease only for January by 0.210C under
RCP8.5 in the midterm, respectively from to baseline period. The increase in mean annual
maximum and minimum temperature under RCP8.5 is larger than that under RCP4.5 scenario
because RCP 8.5 scenario represents a high emission scenario which produces more CO2
concentration than the RCP 4.5 scenario which represents a medium emission scenario (Ba et
al., 2018). These results are in agreement with results from previous studies in the Genale
dawa basin (Tufa and Sarma, 2021; Lemesa, 2017) and Tesfaye et al., 2019) in Awata River
watershed, Genale Dawa basin. Also similar study by (Wakjira et al., 2020) and (Megersa et
al., 2021) in the Finchaa catchment and Upper Wabe Bridge Watershed in Wabe Shebele
River Basin show a consistent increasing trend of maximum and minimum temperature for all
time horizons with a higher rate of increase towards the end of the century.
4.7. Calibration and Validation of SWAT Model

4.7.1. Sensitivity Analysis

Sensitivity analysis was performed on SWAT flow parameters at monthly time step to identify
most sensitive parameters using a global sensitivity analysis of SWAT-CUP at Delo Mena

75
gauging station. All flow parameters were considered to identify which parameters were more
sensitive in Yadot watershed. For this analysis, 21 flow parameters which may affect stream
flow were considered and only 10 parameters were identified to have significant influence in
controlling the stream flow in the watershed (Table 4.10). Maximum canopy storage
(CANMX), SCS runoff curve number (CN2), saturated hydraulic conductivity (SOL_K) and
Soil evaporation compensation factor (ESCO) were found to be the most sensitive parameters
for the stream flow predications. A Study by (Tufa and Sarma, 2021) identified (CN2) and
(SOL_K) most sensitive parameters for runoff estimation in Genale watershed also (Samuel,
2016) identified (CN2) and (CANMX) to be highly sensitive parameters at upper Awash
Basin.

Table 4.10: Sensitive Flow Parameters, their rank and fitted value

Sensitivity P- Min Max Fitted

Parameter Name Rank t-Stat Value Value Value Value
2:V__CANMX.hru 1 -28.82 0.00 0.00 20.50 2.70
1:R__CN2.mgt 2 -23.56 0.00 -0.18 1.70 -0.12
3:R__SOL_K (..).sol 3 12.54 0.00 -0.09 0.21 0.20
4:V__ESCO.hru 4 7.44 0.00 0.89 0.98 0.98
8:V__CH_K2.rte 5 3.92 0.00 30.18 85.08 75.61
6:R__SOL_AWC (..).sol 6 3.57 0.00 -0.10 0.23 -0.01
10:V__GW_DELAY.gw 7 -1.69 0.09 52.19 67.10 61.84
3682.3
9:V__GWQMN.gw 8 0.43 0.67 3379.73 6 3423.16
7:V__ALPHA_BF.gw 9 -0.29 0.77 0.65 0.86 0.81
5:V__RCHRG_DP.gw 10 -0.01 0.99 0.78 0.89 0.88

76
Note: A, indicates add the fitted value to the existing value, V implies replace the existing
value with the fitted value; R indicates multiply the existing value with (1+ the fitted value).

4.7.2. Calibration of SWAT Model

The result of calibration for monthly flow shows that there is a good agreement between the
measured and simulated average monthly flows with Nash-Sutcliffe simulation efficiency
(NSE) of 0.73 and coefficient of determination (R2) of 0.80 as shown below on (Table 4.12)
and (Figure 4.14). Similarly, (Dagnenet et al., 2018) found the correlation coefficient
(R2=0.80) and the Nash – Sutcliffe simulation efficiency (NSE =0.70) which was a very good
agreement between observed and simulated for upper Blue Nile River basin during the
calibration period. Thus, SWAT model can be considered reliable model in simulating stream
flow in Ethiopia condition in different watersheds after calibration.

The measured and simulated average monthly flow for the study area in the calibration period
was obtained to be 9.31m3/s and 9.41m3/s, respectively. This show that average monthly
simulated stream flow is slightly higher than the observed stream flow measured during
stream flow calibration. The model has slightly overestimated stream flow during calibration.
The linear regression of the scattering plot of observed stream flow and simulated stream flow
for calibration is show on (Figure 4.15).

Monthly Areal Rainfall(mm)

Monthly Flow (M3/S)

Areal Rainfall Observed Simulated

100 0
90
80 100
70 200
60
50 300
40
30 400
20 500
10
0 600
88

89

90

91

92

93

94

95

96

97

98

99

00

01

02
n-

n-

n-

n-

n-

n-

n-

n-

n-

n-

n-

n-

n-

n-

n-
Ja

Ja

Ja

Ja

Ja

Ja

Ja

Ja

Ja

Ja

Ja

Ja

Ja

Ja

Ja

Time Series (Month)

Figure 4.14: Calibration result for average monthly stream flow (1988-2002)

77
 Similated Flow (M3/S)
45
40
35
30
f(x) = 1.01835758022257 x − 0.274098583941866
25 R² = 0.796252647317382
20
15
10
5
0
0 5 10 15 20 25 30 35 40

Observed Flow (M3/S)

Figure 4.15: Scatter plots of observed and simulated stream flow for Calibration (1988-2002).
4.7.3. Validation of Model
Validation proves the performance of the model for simulated flows in periods different than
the calibration period without any further adjustment in the calibrated parameters. The
validation was performed for six years, period from 2003 to 2008 shown on (Figure 4.16). The
result of coefficient of determination and Nash-Sutcliff efficiency indicates acceptable
accuracy for the model to predict the watershed response. Based on the model performance
determination coefficient (R2 = 0.83) and Nash-Sutcliff’s simulation efficiency (ENS = 0.77)
assures that the model shows a good performance during validation so as to able to simulate
the stream flow in the study area (Table 4.11). (Wakjira et al., 2020) indicated that the model
performance show a good correlation and agreement between the monthly measured data and
simulated (ENS=0.76 and R2=0.81) flow in Finchaa Watershed for validation period as shown
on The linear regression of the scattering plot of observed stream flow and validation stream
flow is show on (Figure 4.17).

78
 Monthly Areal Rainfall(mm)
Monthly Flow (M3/S)
Areal Rainfall observed simulated
100 0
90
80 100
70 200
60
50 300
40
30 400
20 500
10
0 600
S 03

S 04

S 05

S 06

S 07

S 08
Ja 3

Ja 4

Ja 5

Ja 6

Ja 7

8
M -03

M 04

M -05

M 06

M -07

M 08
-0

-0

-0

-0

-0

-0
-

-
n-

n-

n-
ay

ay

ay

ay

ay

ay
ep

ep

ep

ep

ep

ep
n

n
Ja

Time Series (Month) Fi

Similated Flow (M3/S)

gure 4.16: Validation result for average monthly stream flow (2003-2008)
50
45
40
35
30 f(x) = 1.03449452131243 x + 1.15016997534125
R² = 0.839443410879255
25
20
15
10
5
0
0 5 10 15 20 25 30 35 40
Observed Flow (M3/S)

F
igure 4.17: Scatter plot of observed and simulated stream flow for Validation (2003-2008)
Table 4.11: Model Performance Evaluation Statistics for calibration and validation

Periods Model Performance Evaluation Criteria

R2 NSE % PBIAS RSR P-factors R- factors
Calibration (1988 -2002) 0.80 0.73 16.1 0.52 0.76 0.81
Validation (2003 – 2008) 0.83 0.77 11.2 0.48 0.82 0.79

79
4.8. Independent and Combined effect of Climate and LULCC on Stream Flow

4.8.1. Past Land Use Land Cover Change Impact on Stream flow
One of the most important parts of the study was to assess the impact of past LULC change on
stream flow of the Yadot watershed as summarized in (Table 4.12). The assessment of stream
flow was done in terms of annual, seasonal and monthly basis at outlet of the watershed.

Table 4.12: Mean Annual and Seasonal changes of stream flow results for past LULC change.

Years of Past LULC changes Change detection m3/s (%) change of flow
study 1985 2000 2015 S1 S2 S3 S1 S2 S3

Annual 573.9 576.5 581.14 2.6 4.64 7.24 0.45 0.80 1.26
Wet season 576.5 581.14 576.5 4.47 6.41 10.88 1.10 1.56 2.68
Dry season 410.57 416.98 410.57 -1.87 -1.78 -3.65 -1.11 -1.07 -2.22
Note: S1 Scenarios (1985 to 2000), S2 Scenarios (2000 to 2015) S3 Scenarios (1985 to 2015)

The simulated mean annual stream flows exhibited a higher increase in the second scenarios
4.64m3/s (0.80%) than the first scenarios 2.6m 3/s (0.45%). Unfortunately, the stream flow will
be lower in the first scenarios compared to the second scenarios due to the gradual increase of
grass/range land starting from the year 1985 to 2000. The continued increase of agricultural
and settlement lands and extraction of forest cover and wood land will further increase the
mean annual flow with magnitude of 7.24m 3/s (1.26%) throughout study period from 1985 to
2015.

Seasonal change of stream flows was assessed for the wet (March, April, May, August,
September, October and November) and dry (January, February, June, July and December)
season. As indicated in (Table 4.12) above, the mean wet season flow was increased by
4.47m3/s (1.1%) and 6.41m3/s (1.56%) for the first and second scenarios, respectively. This is
due to rainfall satisfies soil moisture deficit more quickly in the agriculture and settlement than
forest thereby generating more flow in agricultural and settlement. Conversely, the reduction
of vegetation cover and the increase of agriculture and settlement were reduced the dry season
flow by 1.87m3/s (1.11%) and 1.78m3/s (1.07%) for the first and second scenarios,
respectively.

80
 Mean monthly flow (m3/s)
90

75

60

45

30

15

0
JAN FEB MAR APR MAY JUN JUL AUG SEP NOV OCT DEC

Time series (month)

Flow_1985 Flow_2000 Flow_2015

Figure 4.18: Mean monthly stream flow results for the past three land use land cover
Mean monthly stream flow depicted a higher increase in 2015 LULC as compared to 2000 and
1985. This was attributed due to increment of agricultural and settlement and reduction of
forest, scrub/bush land, grass/range land and wood land cover of the watershed. As result of
expansions in agricultural land rainfall infiltrate in to the soil decrease and the soil moisture
shortage more rapidly satisfied under agricultural land than forest thereby producing more
runoff in agricultural land and so, more flow was created in 2000 LULC than 1985 LULC.

Generally, during the study period from 1985 to 2015 mean monthly stream flows increased
by 10.88m3/s for the wet season as a result of agricultural and settlement land expansion by
132.76% and 511.88% and decreases of other LULC separately. These result in agreement
with several studies. For instance (Temesgan et al., 2018) interpreted that, annual stream flow
increase by 300.6mm, 304.5mm and 307.3mm for 1985, 2000 and 2015 LU/LC respectively,
associated with expansion of cultivated land and built-up area and the reduction of forest,
shrub land and grassland in Andassa watershed, Blue Nile Basin, Ethiopia. The study (Moges
et al., 2019) showed the increase in stream flows can be directly attributed to the expansion of
cultivated lands at a cost of the forested vegetation. (Mesfin, 2018) reported expansion of
settlement and agriculture land as result mean monthly stream flow increased by 5.97m 3/s and
0.96m3/s during wet season and dry season flow in shaya catchment Genale Dawa Basin.
Generally (Asmamaw, 2013) reported an increase of cultivated land decline forest and

81
shubland result increased the wet months flow, while the flow during the dry months
decreased.

4.8.2. Future Land Use Land Cover Change Impact on Stream flow
The study in this section mainly focused on analyzing the impact of LULC change on Yadot
watershed stream flow response to future LULC change. The simulated stream flows due to
the future LULC change were compared with past LULC change to estimate the effects of
LULC change on the stream flow.

Table 4.13: Mean Annual and Seasonal change of stream flow results for future LULC change.

Future LULC changes Change detection m3/s % change of flow

Years of 2015 2035 2055 S1 S2 S3 S1 S2 S3
study
Annual 581.14 584.88 588.07 3.74 3.19 6.93 0.64 0.55 1.19
Wet season 416.98 423.1 429.08 6.12 5.98 12.1 1.47 1.41 2.90
Dry season 164.15 161.79 158.99 -2.36 -2.8 -5.16 -1.44 -1.73 -3.14
Note: S1 scenarios (2035 -2015), S2 scenarios (2055 -2035) and S3 scenarios (2015-2055)

Stream flows showed a slight increase in the first scenarios by 3.74m 3/s (0.64%) than the
second scenarios 3.19m3/s (0.55%). Generally stream flows has increased throughout the
future study period with a magnitude of 6.93m3/s (1.19%). Therefore continued increment of
agricultural and settlement and extraction of vegetation cover lead to change the future LULC
of the watershed is expected to further increase the mean annual stream flow.

The amount of mean monthly stream flow was decreased by 2.36m 3/s (1.44%) and 2.8m3/s
(1.73%) for the first and second scenarios, respectively during the dry season. However,
during the wet season mean monthly stream flow was increased by 6.12m 3/s (1.47%) and
5.98m3/s (1.41%) for the first and second scenarios, respectively as shown on (Table 4.13). In
general the future LULC increased mean monthly stream flow by 12.1m 3/s (2.9%) for the
period from 2015 to 2055 during the wet season mainly due to an increase of agricultural and
settlement land by 44.02% and 69.20% which implies that agricultural and settlement land
increased surface runoff. On the other hand, mean monthly stream flows has showed a

82
decreasing trend for the whole study period during dry season with a magnitude of 5.16m 3/s
(3.14%) which reflect that base flow has decreased with an intense agricultural and settlement
expansion.

The stream flow was contributed more in wet months from surface runoff while in dry
months; it was contributed from groundwater. During the study period wet season flow is less
sensitive to LULC change than dry season flow due to ground water contribution during the
dry season was reduced because of less infiltration that largely caused less vegetation cover.
The rate of increment of stream flow was higher for the past LULC compared to future LULC
mainly due to expansion of wood land and grass/range land couple with increasing of
agriculture and settlement.

The result in this study agrees with other previous studies such as (Temesgan, 2018) revealed
that annual and wet season flow, surface runoff and water yield increased while the dry season
flow, groundwater flow, lateral flow and ET have reduced in the periods 1985-2015 due to
LULC change in Upper Blue Nile Basin. The result also shows that for LULC 2030 and 2045
flow expected to increase annually and during wet season, surface runoff and water yield,
while dry season flow, groundwater flow, lateral flow and ET reduce (Rajib and Merwade,
2017) results suggest an increase of 0.5% and 3.5% in the average annual streamflow at the
basin outlet during 2081–2100 for future LULC, respectively. Conversion of cropland, forest,
or grassland to perennial hay/pasture areas would lower surface runoff by 25%, whereas
persistent forest cover in the northern region would cause up to 7% increase in
evapotranspiration.

4.8.3. Impact of Climate Change on Stream flow

The impact of climate change was analyzed taking baseline simulated stream flow against two
future flow in the near and midterm period. The simulated mean annual stream flow may
increase by 7.63% and 5.76% from baseline flow in the near-term under RCP4.5 and RCP8.5
scenarios, respectively. For the midterm mean annual stream flow expected to increase by
8.25% and 6.07%, respectively, from the baseline flow under RCP4.5 and RCP8.5 scenarios.

83
Table 4.14: Mean annual and seasonal change of flow under RCP4.5 and RCP8.5 scenarios
for climate change only from baseline period

Simulated stream flow in (m3/s) % change of stream flow

Scenarios RCP4.5 RCP8.5 RCP4.5 RCP8.5

Near-term

Near-term

Near-term

Near-term
Midterm

Midterm

Midterm

Midterm
Annual 625.46 629.06 614.61 616.40 7.63 8.25 5.76 6.07
Wet season 452.34 456.03 442.97 447.07 8.48 9.36 6.23 7.21
Dry season 173.11 173.03 171.65 169.33 5.46 5.41 4.57 3.16
The seasonal variation of projected stream flow from the baseline period was computed for
wet (March to May and August to November) and dry (January, February, June, July and
December) season. Future change of stream flow for wet and dry season is important to
understand hydrological impact of climate change. The wet season mean stream flow of the
watershed expected to increase by 8.48% and 6.23% from the baseline flow in the near-term
under RCP4.5 and RCP8.5 scenarios, respectively. For the midterm mean wet season stream
flow may increase by 9.36% and 7.21% from the baseline stream flow under RCP4.5 and
RCP8.5scenarios, respectively (Table 4.15). Dry season mean stream flows is projected to
increase in the watershed by 5.46% and 4.57% from baseline flow in the near-term under
RCP4.5 and RCP8.5scenarios, respectively. Similarly midterm the mean dry season stream
flow expected to increase by 5.41% and 3.16% from the baseline stream flow under RCP4.5
and RCP8.5scenarios, respectively.

The mean monthly percentage change of stream flow in the near and midterm for RCP4.5 and
RCP8.5 scenarios from the baseline stream flow are presented in (Figure 4.19). The mean
monthly stream flow may expected to increase for all months except for the month of
February, July, August and September showed a decrease in the near and midterm under
RCP4.5 and RCP8.5 scenarios from the baseline period, respectively.

84
 140 140
120 120
100 100

Average Monthly Flow (m3/s)

Average Monthly Flow (m3/s)

80 80
60 60
40 40
20 20
0 0
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC

Time Series (Month)

Time Series (Month)

Baseline NEAR RCP4.5 MID RCP4.5 Baseline NEAR RCP8.5 MID RCP8.5
F
igure 4.19: Mean monthly flow for near and midterm under RCP4.5 and RCP8.5 scenarios
from the baseline period.
The maximum increase in mean monthly stream flown expected in May with 36.6m 3/s
(40.75%) and 34.16m3/s (38.02%) in the near and midterm, while highly decline in August by
28.8m3/s (64.21%) and 28.16m3/s (62.78%) for near and midterm under RCP4.5 from the
baseline flow, respectively.

60 60

40 40
%Change of Stream Flow
%Change of Stream Flow

20 20

0 0
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
-20 -20

-40 -40

-60 -60

-80 -80
Time Series (Month) Time Series (Month)

NEAR RCP4.5 MID RCP4.5 NEAR RCP8.5 MID RCP8.5

Figure 4.20: Percentage change of flow for near and midterm future under RCP4.5 and
RCP8.5 scenarios compared to baseline period.
Mean monthly stream flow showed highest increase for the month of May by 31.16m3/s
(34.68%) and 30.75m3/s (34.22%), while maximum reduction occurred in August by
29.61m3/s (66%) and 27.53m3/s (61.38%) from the baseline flow in the near and midterm

85
under RCP8.5 scenarios, respectively. The high flow months in the baseline period such as
August and September will be highly impacted in the near and midterm for both RCP 4.5 and
RCP 8.5 scenarios as shown on (Figure 4.20). In general overall trend shows that the annually,
seasonal and monthly change of stream flow expected to increase in all future of Yadot
watershed under RCP4.5 and RCP8.5 scenarios.

The projected increase in stream flow into the watershed is in agreement with studies (Tufa
and Sarma, 2021) reported that mean annual stream flow consistently increased with the
predicted changes in rainfall and temperature patterns in the future period 2022–2080 under
the RCP 4.5 and RCP 8.5 emission scenarios of Genale watershed. (Tesfaye et al, 2019)
summaries increase in average total annual, seasonal and monthly flow volume is observed for
the period (2018 to 2077) corresponding to increase in mean annual, seasonal and monthly
precipitation during future scenario of Awata river watershed, Genale Dawa Basin. Similarly
(Anwar et al., 2016) conclude that Stream flow projections for future time periods showed that
mean annual stream flow increase by 7.1, 9.7, and 10.1 % at 2020s, 2050s, and 2080s,
respectively, from the baseline period for A2 scenario, whereas for B2 scenario, it will be
expected to increase by 6.8, 7.9, and 6.4 % for 2020s, 2040s, and 2080s, respectively in the
Upper Gilgel Abay Watershed, Blue Nile basin. Therefore, the fluctuation mean stream flow
for annual, wet and dry month will result from climate change in the Bale highlands not only
affects the livelihood of people in the Bale zone but also all inhabitants at the downstream
areas whose life depends on the flow of yadot River.

4.8.4. Combined Impact of Climate and LULC Change on Stream flow

To determine the combined impacts of climate and LULC change on steam flow, four
simulations were carried out by using a combination of two futures LULC (2035 for near term
and 2055 for midterm) with two RCP4.5 and RCP8.5 scenarios. The combined impacts
climate and future land use land cover change on stream flow in the study area were simulated
using the calibrated SWAT model the results of which are presented on (Table 4.15).

The simulated mean annual stream flow showed increasing trend under both climate scenarios
in the future projection periods. The simulated mean annual stream flow showed an increasing

86
by 8.13% and 6.26% in the near-term from baseline period under RCP4.5 and RCP8.5
scenarios, respectively. Compared to RCP4.5 in the near-term RCP8.5 showed slight decline
of mean annual stream flow mainly due to annual and seasonal decline of precipitation from
baseline period with an increase in temperature over the watershed.

Table 4.15: Mean Annual and Seasonal change of flow under RCP4.5 and RCP8.5 scenarios
for combined climate and LULC change from baseline period.

Stream flow in (m3/s) % change of stream flow

Scenarios RCP4.5 RCP8.5 RCP4.5 RCP8.5
Near-term

Near-term

Near-term

Near-term
Midterm

Midterm

Midterm

Midterm
Annual 628.36 633.23 617.52 620.47 8.13 8.96 6.26 6.77
Wet season 453.34 451.98 443.92 443.07 8.72 8.39 6.46 6.25
Dry season 175.02 181.24 173.60 177.40 6.62 10.42 5.76 8.07
Note: Near-term (2021-2050) and midterm (2051 -2080)

In the midterm projected mean annual stream flow increased by 8.96% and 6.77% in the
midterm under RCP4.5 and RCP8.5 scenarios from baseline flow, respectively (Table 4.16).
The change in stream flow for impact of combined climate and LULC change is consistent
with climate change alone. The highest the rainfall under RCP8.5 in the midterm the lower the
simulated mean annual stream flow compared to RCP4.5 in the same time period mainly due
to high temperature and increased evapo-transpiration in the midterm. If change in land use
land cover, were held constant for both scenarios. There will be higher rates of evaporation in
Ethiopia due to warming over the country. Such changes are expected to impact the economy
of East African countries, including Ethiopia (Change 2014). Therefore stream flow was
sensitive to change in the temperature than rainfall and LULC change for the midterm in the
Yadot watershed.

The seasonal variation of stream flow for the projected climate parameter change from the
baseline period was computed for wet and dry season (Table 4.15). The mean wet season

87
stream flow may increase by 8.72% and 6.46%, while the dry season flow may increase by
6.62% and 5.76% from baseline period for near-term under RCP4.5 and RCP8.5 scenarios,
respectively. For the midterm mean wet season flow will increase by 8.39% and 6.25%, while
the dry season flow will be expected to increase by 8.07% and 10.42% from baseline period
under RCP 4.5 and RCP 8.5 scenarios, respectively. The combination of LULC and climate
change scenario will have an additive effect on stream flow simulation. Therefore, simulation
results showed that if all other variables are held constant, an increase in river flow may be
expected in the future as a consequence of climate and LULC change, while the impacts
associated should be effectively planned for and mitigated. Therefore mean wet and dry
season stream flow may increase for the coming century under both RCP from baseline
period.

Mean monthly change of steam flow indicate there may be an increase for the months of
(January, March, April, May, June, October, November and December), while decrease for the
month of (February, July, August and September) under both RCP from baseline period for
near and midterm, respectively as shown on (Figure 4.21).

140 140
120 120
100 100
80
Average Monthly Flow (m3/s)

80
Average Monthly Flow (m3/s)

60 60
40 40
20
20
0
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC 0
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC

Time Series (Month) Time Series (Month)

Baseline NEAR RCP4.5 NEAR RCP8.5 Baseline MID RCP4.5 MID RCP8.5
F
igure 4.21: Mean monthly stream flow for near and midterm under RCP4.5 and RCP8.5
scenarios compared to baseline period.
The maximum mean monthly change of flow increment was observed in May by 35.39m 3/s
(39.4%) and 31.77m3/s(35.36%), while minimum mean monthly steam flow increment was in

88
December and March by 5.87m3/s (15.81%) and 2.3m3/s (11%) for near-term under RCP4.5
and RCP8.5 from baseline period. Mean monthly change of flow showed maximum reduction
in August by 28.66m3/s (63.9%) and 27.12m3/s (60.46%), whereas the minimum decline of
flow in February by 1.53m3/s (12.3%) and 2.3m3/s (18.45%) for near-term from baseline
period under RCP4.5 and RCP 8.5, respectively.

The largest percentage change of mean monthly flow increment observed in May by 34.39m 3/s
(38.28%) and 31.27m3/s (34.80%), while the future mean monthly change of flow showed
smallest nominal increment in March by 4.19m3/s (20.09%) and 2.3m3/s (11.03%), from
baseline period in the midterm under RCP4.5 and RCP8.5, respectively (Figure 4.22).

60 80

40 60
%Change of Stream Flow
%Change of Stream Flow

40
20
20
0
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC 0
-20 JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
-20
-40
-40
-60 -60
-80 -80
Time Series (Month) Time Series (Month)

NEAR RCP4.5 NEAR RCP8.5 MID RCP4.5 MID RCP8.5

F
igure 4.22: Percentage change of flow for near-term and midterm under RCP4.5 and RCP8.5
scenarios compared to baseline period.
The simulated mean monthly steam flow change showed significant decline in August by
27.7m3/s (61.76%) 27.12m3/s (60.46%), whereas the minimum reduction in mean monthly
flow occurred in February by 0.6m3/s (4.82%) and 1.3m3/s (10.42%) in the midterm from
baseline period under RCP4.5 and RCP8.5, respectively. Comparing the impacts of climate
change and LULC change on stream flow, it is clear that climate change impact was dominant
and its impact increase stream flow compared to baseline period. Stream flow is more
sensitive to climate change than to changes in LULC and thus the changes in stream flow
under future conditions of climate and LULC change are remarkably consistent with those in
which future climate changed and secondary role with change in LULC. Changes in future

89
LULC produced small changes in stream flow and combined scenario predicted larger changes
in stream flow than which considered under only LULC change and only climate change from
baseline flow. Therefore results show that the combined climate and LULC change have a
more noticeable effect on stream flow in comparison with the baseline period, the stream flow
gradually increased in the future in general.

The results of this study is consistent with other regional and local climate change impact
assessment studies by (Birhan et al., 2021) the increasing trend to of stream flows in relation
to temperature and precipitation increase in the basin by the end of the 21 st century in the Lake
Tana Basin, upper Blue Nile River Basin. A similar study in Kesem sub-basin of the Awash
River basin, showed an increase in mean annual stream flow by 14.5 and 19.1% for RCP 4.5
and by 4.7 and 6.9% for RCP 8.5 scenarios parallel to rainfall and temperature increases from
the 2050s to 2080 period (Negash et al., 2020). (Dagnenet et al., 2018) reported that
precipitation increase by up to 25% in the Upper Blue Nile River Basin which result in
projected increase of stream flow and lateral flow by up to 12.85% (9.9%), 28.5% (20.03%)
and 26.4% (29.12%) under LULC, climate and combined climate and LULC change
scenarios, respectively. Study (Babur et al., 2016) report from Pakistan reported an increase in
mean annual flow under RCP 4.5 and RCP 8.5 scenarios between the years 2011–2040, 2041–
2070, and 2071–2100. Therefore, soil and water conservation activities should be adopted by
the community as well as the water harvesting structure should be properly designed and
applied to the watershed to compensate for this fluctuation of flow in the Yadot River. Also
larger increment of stream flow from the baseline scenario may cause flooding to flood plain
zones. It is crucial to consider and implement integrated water resources management to meet
the alarmingly increasing water demand due to high population growth rate.

90
5. SUMMARY CONCLUSIONS AND RECOMMENDATION

5.1. Summary and Conclusions

Climate and land use land cover changes are the two inseparable linked global environmental
challenges the world faces today (IPCC, 2019). In this study the potential impact of climate
and land use land cover change on stream flow of Yadot watershed analyzed by considering
past and future LULC change and climate change indicator parameters (precipitation and
temperature) as an input of the SWAT model. CA-Markov model was used to predict future
LULC change for 2035 and 2055 using IDRISI software. In order to detect and analyze
changes in LULC for the study periods of 1985, 2000 and 2015 GIS and Remote Sensing
techniques were applied to generate LULC maps through a maximum likelihood supervised
image classification algorithm. Over the past three decades, the predominant patterns of LULC
change were the transitional conversions among agricultural land, settlement, grass/range land,
wood land, forest land and Scrub/bush land.

The dominant changes were decline of forest and scrub/bush land coverage and an increase of
agriculture and Settlement with total human population increment. The LULC change results
showed that throughout the study period, agricultural and settlement land significantly
increased from 6.11% and 0.12% to 20.47% and 1.24%, respectively, while scrub/bush land
and forest continuously decline from 65.5% and 7.45% to 45.22% and 6.35%, respectively.
Conversely on average grass/range land and wood land shows decline in the past from 7.54%
and 13.29% to 7.3% and 9.17% and expected to increase from 2015 to 235 to 9.66% and
16.05%, respectively, while it will decline to 9.51% and 14.21% from 2035 to 2055 in the
total coverage of the watershed, respectively.

The ensemble mean of three RCMs output in the coordinated regional climate downscaling
experiment (CORDEX)-RCM shows parallel daily precipitation and temperature increase in
the two future periods under RCP 4.5 and RCP 8.5 but varies on monthly basis. Except in the
near-term under RCP8.5 precipitation decrease compared to baseline and highest increment of
precipitation observed in the midterm under RCP8.5 scenarios. Precipitation may increase in
some months and decrease in other months for both RCP4.5 and RCP8.5 scenario in all future.

91
The mean annual maximum temperatures projected to increase by 0.65 0C and 1.050C, while
the mean annual minimum temperature would increase by 0.48 0C and 0.850C under RCP4.5
scenario in the near and midterm, respectively, while mean annual maximum temperature
simulated to increase by 0.710C and 20C, while mean annual minimum temperature was
increase by 0.560Cand 1.80C under RCP 8.5 scenario in the near and midterm from baseline
period, respectively. The overall study result shows the projected maximum and minimum
temperature under RCP8.5 is warmer than RCP4.5. Compared with the historical period, the
annual mean temperature would face a continuously increasing trend under both scenarios
from baseline. In general annual precipitation and mean annual temperature would increase to
varying degrees in the future.

The result of the dynamically downscaling Model outputs for future scenario on a monthly and
seasonal basis indicates that precipitation does not show a systematic increase or decrease.
However annual and seasonal flow shows an increasing trend under RCP4.5 and RCP8.5
scenarios compared to the baseline period. The mean annual stream flow projected to increase
by 7.63% (8.13%) and 5.76% (6.26%) for near-term (2021-2050), whereas for the midterm
(2051 – 2080) model shows an increase by 5.76% (6.26) and 6.07% (6.72%) at the outlet of
the watershed for both climate only and combined climate and LULC change from baseline
period for both RCP4.5 and RCP8.5 scenarios, respectively. Predicted seasonal flow shows an
increase for dry and wet seasons compared to baseline period under RCP4.5 and RCP8.5
scenarios in all future. The dry season flow show a highest increasing trend for the midterm
under both scenarios compared to wet season flow. Mean monthly change of stream flow
clearly shows the season which may be impacted highly with climate change

Generally, the result of future stream flow projection suggests that the combined scenarios
were slightly higher than the climate change alone. However, increase in future stream flow
primarily driven by future climate change, whereas the changes in LULC play a secondary
role, but highest the rainfall under RCP8.5 in the midterm the lower the simulated mean
annual stream flow compared to RCP4.5 in the same time period mainly due to high
temperature and increased evapo-transpiration in the midterm.

92
5.2. Recommendation
Depending on the results of the study the following recommendation:-

The watershed water management system should be following the future trends of rainfall
peaks as the temporal shift in peak rainfall showed a direct impact on the flow of Yadot
river watershed Therefore, soil and water conservation activities should be adopted by the
community as well as the water harvesting structure should be properly designed for the
effective planning and management practices of soil and water resources applied to the
watershed to compensate the fluctuation of flow in the Yadot River.
Water play significant role for small & large scale irrigation, therefore mitigating adverse
impacts of climate and LULC change is important. This study will then be meaningful to
society and so that they can either be well adapted to the forthcoming climatic condition or
mitigate the adverse impacts of changing climate and land use land cover change.
For future, it is better to further analysis of other parameters in addition to climate and land
use land cover change such as land use land cover change alone, population growth rate
and water utilization efficiency to find out the characteristics’ changes of on hydrology
that might influenced by characteristics of these factors in Genale subbasin. It also
important to study with the same title using more than three RCMs output.

93
6. REFERENCE
Abbas, N, S.A Wasimia and N Al-Ansari, 2016. Assessment of climate change impacts on
water resources of Al-Adhaim, Iraq using SWAT model. Engineering, 8, pp.716-732.

Abdi Boru, Desalegn Chemeda and Ekasit Kositsakulchai, (2014). "“Modelling the effects of
land use changes and management practices on Runoff and Sediment Yields in Fincha
Watershed, Blue Nile.”." OIDA International Journal of Sustainable Development
07:11.

Abdulkareem, J.H., B. Pradhan, Sulaiman and N.R Jamil, 2018. Review of studies on
hydrological modelling in Malaysia. Modeling Earth Systems and Environment, 4(4),
pp.1577-1605.

Abebe Tariko, 2020. Modeling Impact of Land Use Dynamics on Hydrology and
Sedimentation of Megech Dam Watershed, Ethiopia. The Scientific World
Journal, 2020.

Aboelnour, M., M.W. Gitau and B.A Engel, 2019. Hydrologic response in an urban watershed
as affected by climate and land-use change. Water, 11(8), p.1603.

Adane Mezgebu and Getachew Workineh, 2017. Changes and drivers of afro-alpine forest
ecosystem: future trajectories and management strategies in Bale eco-region,
Ethiopia. Ecological Processes, 6(1), pp.1-13.

Rajib. A, and Merwade.V, ,2017. Hydrologic response to future land use change in the Upper
Mississippi River Basin by the end of 21st century. Hydrological Processes. 2017 Oct
15;31(21):3645-61.

Afera Halefom, Asirat Teshome, Ermais Sisay, D.Khare, Mihret Dananto, L. Singh and
Dessalew Tadesse, 2018. Applications of Remote Sensing and GIS in Land Use/Land
Cover Change Detection: A Case Study of Woreta Zuria Watershed, Ethiopia. Applied
Research Journal of Geographic Information System, 1(1), pp.1-9.

94
Ahn, K.H. and V Merwade 2014. Quantifying the relative impact of climate and human
activities on streamflow. Journal of Hydrology, 515, pp.257-266.

Aich, V., S Liersch, T.Vetter, S.Huang, J .Tecklenburg, P.Hoffmann, H.Koch, S. Fournet V

Krysanova, E.N Müller and F.F Hattermann, 2014. Comparing impacts of climate
change on streamflow in four large African river basins. Hydrology and Earth System
Sciences, 18(4), pp.1305-1321.

Al-sharif, A., Pradhan, B. (2013). Monitoring and predicting land use change in Tripoli
Metropolitan City using an integrated Markov chain and cellular automata models in
GIS. Arab J Geosci. 7: 4291-4301.

Amare Sewnet, 2013). Retrospective Analysis of Land Cover and Use Dynamics in
GilgelAbbay Watershed by Using GIS and Remote Sensing Techniques, Northwestern
Ethiopia.International Journal of Geosciences, 4:1003-1008.

Amisigo, B.A., Logah, F.Y., Obuobie, E. and Kankam-Yeboah, K., 2018. Impacts of Climate
Change on Stream Inflows into the Volta Lake. Ghana Journal of Science, 58, pp.23-3

Anand, J., Gosain, A. K., Khosa, R. 2018. Prediction of land use changes based on Land
Change Modeler and attribution of changes in the water balance of Ganga basin to land
use change using the SWAT model. Science of the Total Environment, 644, 503-519.

Andargachew melke and Fantahun Abegaz, 2017. Impact of climate change on hydrological
responses of Gumara catchment, in the Lake Tana Basin-Upper Blue Nile Basin of
Ethiopia. International Journal of Water Resources and Environmental
Engineering, 9(1), pp.8-21.

Andualem, T., Belya, G. and Guadie, A. 2018. Land Use Change Detection Using Remote
Sensing Technology. . J Earth Sci Clim Change 9, 488-496.

Anwar. A, Seifu.A, Essayas .K, Abeyou .W. Tewodros T. Shimelis .B and Assefa M, 2016.
Climate change impact on stream flow in the upper Gilgel Abay Catchment, Blue Nile
Basin, Ethiopia. In Landscape Dynamics, Soils and Hydrological Processes in Varied
Climates (pp. 645-673). Springer, Cham.

95
Arnold, J.G., Moriasi, D.N., Gassman, P.W., Abbaspour, K.C., White, M.J., Srinivasan, R.,
Santhi, C., Harmel, R.D., Van Griensven, A., Van Liew, M.W. and Kannan, N., 2012.
SWAT: Model use, calibration, and validation. Transactions of the ASABE, 55(4),
pp.1491-1508.

Arsanjani, J. J., Helbich, M., Kainz, W., Boloorani, A. D. (2013): Integration of logistic
regression, Markov chain and cellular automata models to simulate urban expansion.
International Journal of Applied Earth Observation and Geoinformation 21: 265-275.

Asirat Teshome, 2018. Evaluating the Dynamics of Land Use/Land Cover Change Using GIS
and Remote Sensing Data in Case of Yewoll Watershed, Blue Nile Basin, Ethiopia.

Asirat, Teshome, 2018. Evaluating the Dynamics of Land Use/Land Cover Change Using GIS
and Remote Sensing Data in case of Yewoll Watershed, Blue Nile Basin, Ethiopia. .
Applied Research Journalof Geographic Information System. , 1, 10-24.

Asmamaw Adamu, 2013. Assessing the impacts of land use and land cover change on
hydrology of watershed: a case study on Gigel-Abbay Watershed, Lake Tana Basin,
Ethiopia (Doctoral dissertation).

Azari, M., Moradi, H.R., Saghafian, B. and Faramarzi, M., 2016. Climate change impacts on
streamflow and sediment yield in the North of Iran. Hydrological Sciences Journal,
61(1), pp.123-133.

Ba, W., Du, P., Liu, T., Bao, A., Luo, M., Hassan, M. and Qin, C., 2018. Simulating
hydrological responses to climate change using dynamic and statistical downscaling
methods: a case study in the Kaidu River Basin, Xinjiang, China. Journal of Arid
Land, 10(6), pp.905-920.

Babur, M., Babel, M.S., Shrestha, S., Kawasaki, A. and Tripathi, N.K., 2016. The Impact of
Climate Change on Reservoir Inflows Using Multi Climate-Model under RCPs’
Including Extreme Events—A Case of Mangla Dam, Pakistan.

96
Belay Simane, Hunachew Beyene, Wakgari Deressa, Abera Kumie, Kiros Berhane and
Jonathan Samet (2017). Review of Climate Change and Health in Ethiopia: Status and
Gap Analysis. Ethiopia Journal of Health Development, 30(1), 1–23.

Belkele Megersa, Andre Markeman, Angassa and A.V. Zárate., 2014. The role of livestock
diversification in ensuring household food security under a changing climate in
Borana, Ethiopia. Food Security, 6(1), pp.15-28.

Binyam Alemu, Efrem Garedew, Zewdu Eshetu and Habtemariam Kassa, 2015. Land use and
land cover changes and associated driving forces in north western lowlands of
Ethiopia. Int Res J Agric Sci Soil Sci. 5: 28–44.

Bireda Alemayehu, 2015. GIS and remote sensing based land use/land cover change detection
and prediction in Fagita Lekoma Woreda, Awi Zone, North Western Ethiopia. Addis
Ababa, Ethiopia: Addis Ababa University.

Birhan Asmame and Assefa Abegaz, 2018. Land use/land cover changes and their
environmental implications in the Gelana sub-watershed of Northern highlands of
Ethiopia. Environmental Systems Research, 6(1), pp.1-12.

Birhan Getachew, B.R Manjunatha, and H.G Bhat, 2021. Modeling projected impacts of
climate and land use/land cover changes on hydrological responses in the Lake Tana
Basin, upper Blue Nile River Basin, Ethiopia. Journal of Hydrology, 595, p.125974.

Birhnu Liyew, Atsushi Tsunekawa Nigussie Haregeweyn, Derege Tsegaye, Enyew Adgo,
Mitsuru Tsubo, Tsugiyiki Masunaga, Ayele Almaw, Dagenrnt Sultan Mesenbet
Yibeltal, 2019. Exploring land use/land cover changes, drivers and their implications in
contrasting agro-ecological environments of Ethiopia. Land Use Policy, 87, p.104052.

Bjørnæs, F., 2015. Som rekrutteringsrådgiver er det mitt største ansvar å komme bak Halo-
effekten (Master's thesis, NTNU).

Chang, J., Ciais, P., Viovy, N., Vuichard, N., Sultan, B. and Soussana, J.F., 2015. The
greenhouse gas balance of European grasslands. Global change biology, 21(10),
pp.3748-3761.

97
Charles, J.P., Iwema, T., Epa, V.C., Takaki, K., Rynes, J. and Jindra, M., 2011. Ligand-
binding properties of a juvenile hormone receptor, Methoprene-tolerant. Proceedings
of the National Academy of Sciences, 108(52), pp.21128-21133.

Chen, J., Brissette, F.P., Leconte, R. and Caron, A., 2012. A versatile weather generator for
daily precipitation and temperature. Transactions of the ASABE, 55(3), pp.895-906.

Clark Labs. (2012). Idrisi Selva help system. Clark University, USA.

Clerici, N., Cote-Navarro, F., Escobedo, F.J., Rubiano, K. and Villegas, J.C., 2019. Spatio-
temporal and cumulative effects of land use-land cover and climate change on two
ecosystem services in the Colombian Andes. Science of the Total Environment, 685,
pp.1181-1192.

Dagnenet Fenta, Z. Duan, T. Rientjes, and Markus Disse., 2018. Analysis of combined and
isolated effects of land-use and land-cover changes and climate change on the upper
Blue Nile River basin's streamflow. Hydrology and earth system sciences, 22(12),
pp.6187-6207.

Desalegn, T., Cruz, F., Kindu, M., Turrión, M.B. and Gonzalo, J., 2014. Land-use/land-cover
(LULC) change and socioeconomic conditions of local community in the central
highlands of Ethiopia. International Journal of Sustainable Development & World
Ecology, 21(5), pp.406-413.

Devkota, L.P. and Gyawali, D.R., 2015. Impacts of climate change on hydrological regime
and water resources management of the Koshi River Basin, Nepal. Journal of
Hydrology: Regional Studies, 4, pp.502-515.

Di Gregorio, A., 2016. Land covers classification system: classification concepts. Software
version 3.

Domínguez, M., Romera, R., Sánchez, E., Fita, L., Fernández, J., Jiménez-Guerrero, P.,
Montávez, J.P., Cabos, W.D., Liguori, G. and Gaertner, M.Á., 2013. Present-climate
precipitation and temperature extremes over Spain from a set of high resolution
RCMs. Climate research, 58(2), pp.149-164.

98
Eastman, J. (2012). IDRISI Selva manual, version 17. Clark University. p322.

Ebrahim Esa and Mohamed Assen, 2018. Land use/cover dynamics and its drivers in Gelda
catchment, Lake Tana watershed, Ethiopia. Environmental Systems Research, 6(1),
pp.1-13.

Emerta, B. and Aragie, A. (2013). Climate change, growth, and poverty in Ethiopia. TEXAS
UNIV AT AUSTIN.

Ermias Teferi, Woldeamlak Bewket, S. Uhlenbrook, and J. Wenninger, 2013. Understanding

recent land use and land cover dynamics in the source region of the Upper Blue Nile,
Ethiopia: Spatially explicit statistical modeling of systematic transitions. Agriculture,
ecosystems & environment, 165, pp.98-117.

Evans, J.P., 2011. Resilience, ecology and adaptation in the experimental city. Transactions of
the institute of British Geographers, 36(2), pp.223-237.

Fasil, V.K. and Karmakar, S., 2012, July. Modeling and simulation based study for on-line
detection of partial discharge of solid dielectric. In 2012 IEEE 10th International
Conference on the Properties and Applications of Dielectric Materials (pp. 1-5). IEEE.

Gambo, J., Shafri, H.M., Shaharum, N.S.N., Abidin, F.A.Z. and Rahman, M.T.A., 2018.
Monitoring and predicting land use-land cover (LULC) changes within and around
Krau wildlife reserve (KWR) protected area in Malaysia using multi-temporal Landsat
data. Geoplanning: Journal of Geomatics and Planning, 5(1), pp.17-34.

Gebre Hadgu, kindie Tesfaye and Girma Mamo Belay Kasa, 2013. Trend and variability of
rainfall in Tigray, northern Ethiopia: analysis of meteorological data and farmers'
perception. Academia Journal of Agricultural Research, 1(6), pp.088-100.

Gebre Hadgu, kindie Tesfaye and Girma Mamo, 2015. Analysis of climate change in Northern
Ethiopia: implications for agricultural production. Theoretical and Applied
Climatology, 121(3), pp.733-747.

99
Goitom Gebreluel, 2014. Ethiopia's Grand Renaissance Dam: ending Africa's oldest
geopolitical rivalry?. The Washington Quarterly, 37(2), pp.25-37.

Gudeta W. Sileshi, 2016. The magnitude and spatial extent of influence of Faidherbia albida
trees on soil properties and primary productivity in drylands. Journal of Arid
Environments, 132, pp.1-14.

Guo, Y., Fang, G., Wen, X., Lei, X., Yuan, Y. and Fu, X., 2019. Hydrological responses and
adaptive potential of cascaded reservoirs under climate change in Yuan River
Basin. Hydrology Research, 50(1), pp.358-378.

Haile E Getachew and Assefa M Melesse, 2012. The impact of land use change on the
hydrology of the Angereb Watershed, Ethiopia. International Journal of Water
Sciences, 1(6).

Setegn, S.G., Srinivasan, R. and Dargahi, B., 2008. Hydrological modelling in the Lake Tana
Basin, Ethiopia using SWAT model. The Open Hydrology Journal, 2(1).

Halmy, M. W. A., Gessler, P. E., Hicke, J. A., Salem, B. B. (2015): Land use/land cover
change detection and prediction in the north-western coastal desert of Egypt using
Markov CA. – Applied Geography 63: 101-112.

Hartmann, D.L., Tank, A.M.K., Rusticucci, M., Alexander, L.V., Brönnimann, S., Charabi,
Y.A.R., Dentener, F.J., Dlugokencky, E.J., Easterling, D.R., Kaplan, A. and Soden,
B.J., 2013. Observations: atmosphere and surface. In Climate change 2013 the physical
science basis: Working group I contribution to the fifth assessment report of the
intergovernmental panel on climate change (pp. 159-254). Cambridge University.

Headey, D., Dereje, M. and Taffesse, A.S., 2014. Land constraints and agricultural
intensification in Ethiopia: A village-level analysis of high-potential areas. Food
Policy, 48, pp.129-141.

Hernández-Díaz, L., Laprise, R., Sushama, L., Martynov, A., Winger, K. and Dugas, B., 2013.
Climate simulation over CORDEX Africa domain using the fifth-generation Canadian
Regional Climate Model (CRCM5). Climate Dynamics, 40(5-6), pp.1415-1433.

100
Hussen Seid, P Omond, S. Jain, C. Lennard, B. Hewitson, L. Chang'a, J.L. Awange, Dosio, A.,
P. Ketiem, G. Nikulin, and H.J Panitz, 2013. Assessment of the performance of
CORDEX regional climate models in simulating East African rainfall. Journal of
Climate, 26(21), pp.8453-8475.

Hyandye, C.B., Worqul, A., Martz, L.W. and Muzuka, A.N., 2018. The impact of future
climate and land use/cover change on water resources in the Ndembera watershed and
their mitigation and adaptation strategies. Environmental Systems Research, 7(1),
pp.1-24.

IPCC Climate Change, 2013. The physical science basis. Contribution of working group I to
the fifth assessment report of the intergovernmental panel on climate change, pp.159-
254.

IPCC. (2019). Climate Change and Land: IPCC Special Report on Climate Change,
Desertification, Land Degradation, Sustainable Land Management, Food Security, and
Greenhouse gas fluxes in Terrestrial Ecosystems. 1–1542.

IPCC-TGICA, A., 2007. General guidelines on the use of scenario data for climate impact and
adaptation assessment (p. 66). version 2. Tech. rep.

Jones, C., Giorgi, F. and Asrar, G., 2011. The Coordinated Regional Downscaling
Experiment: CORDEX–an international downscaling link to CMIP5. CLIVAR
exchanges, 16(2), pp.34-40.

Kassie, B.T., Rötter, R.P., Hengsdijk, H., Asseng, S., Van Ittersum, M.K., Kahiluoto, H. and
Van Keulen, H., 2014. Climate variability and change in the Central Rift Valley of
Ethiopia: challenges for rainfed crop production. The Journal of Agricultural
Science, 152(1), pp.58-74.

Kidane Welde, 2016. Identification and prioritization of subwatersheds for land and water
management in Tekeze dam watershed, Northern Ethiopia. International Soil and
Water Conservation Research, 4(1), pp.30-38.

101
Kinati Chimdessa, S. Quraishi, Asfaw Kebede Tena Alamirew, 2019. Effect of land use land
cover and climate change on river flow and soil loss in Didessa River Basin, South
West Blue Nile, Ethiopia. Hydrology, 6(1), p.2.

Koch, H., Biewald, A., Liersch, S., de Azevedo, J.R.G., da Silva, G.S., Kölling, K., Fischer,
P., Koch, R. and Hattermann, F.F., 2015. Scenarios of climate and land-use change,
water demand and water availability for the São Francisco River Basin. Brazilian
Journal of Environmental Sciences (Online), (36), pp.96-114.

Koomen, E., Borsboom-van Beurden, J. (2011): Land-use modelling in planning practice.

Springer, XVI-214.

Leal Filho, W., Azeiteiro, U.M. and Setti, A.F.F., 2021. Challenges in Sustainable Land Use
Management. In Sustainability in Natural Resources Management and Land
Planning (pp. 553-556). Springer, Cham.

Lemesa Muleta, 2017. Impact of climate change on water availability in Genale River Bub-
basin, southeastern Ethiopia (Doctoral dissertation, Haramaya University).

Lennard, C.J., Nikulin, G., Dosio, A. and Moufouma-Okia, W., 2018. On the need for regional
climate information over Africa under varying levels of global
warming. Environmental Research Letters, 13(6), p.060401.

Li, J., Zhang, X.Y. and Yang, Y.Z., 2012. SWAT model of runoff study under different land
use land cover scenarios in source region of the Yangtze River. Research of Soil and
Water Conservation, 19(3), pp.119-124.

Li, J., Zheng, X., Zhang, C. and Chen, Y., 2018. Impact of land-use and land-cover change on
meteorology in the Beijing–Tianjin–Hebei Region from 1990 to
2010. Sustainability, 10(1), p.176.

Li, S., Jin, B., Wei, X., Jiang, Y., Wang, J. (2015). Using CA-Markov model to model the
spatiotemporal change of land use/cover in Fuxian Lake for decision support.
International workshop on spatiotemporal computing, 13–15 July 2015, Fairfax,
Virginia,

102
Liersch, S., Tecklenburg, J., Rust, H., Dobler, A., Fischer, M., Kruschke, T., Koch, H. and
Hattermann, F.F., 2018. Are we using the right fuel to drive hydrological models? A
climate impact study in the Upper Blue Nile. Hydrology and Earth System
Sciences, 22(4), pp.2163-2185.

Masui, T., Matsumoto, K., Hijioka, Y., Kinoshita, T., Nozawa, T., Ishiwatari, S., Kato, E.,
Shukla, P.R., Yamagata, Y. and Kainuma, M., 2011. An emission pathway for
stabilization at 6 Wm− 2 radiative forcing. Climatic change, 109(1), pp.59-76.

Megersa Adugna Nura Boru and Abebe Debele, 2021. Modeling Climate Change Impact on
the Streamflow in the Upper Wabe Bridge Watershed in Wabe Shebele River Basin,
Ethiopia. International Journal of River Basin Management, (just-accepted), pp.1-46.

Megersa Kebede and A. Chakravarti, 2017. Sediment Yield Assessment and Mitigation
Measures in Finchaa Watershed, Ethiopia. Int. J. Eng. Res. Technol, 6, pp.220-226.

Mesfin Reta, 2018. Impact of land use and land cover change on stream flow using mike she:
case study of Shaya catchment, Ethiopia.

Mesgana Seyoum, Getu Fana, T.Y Gan, Semu Ayalew and H. Koivusalo, 2017. Potential
impact of climate change on streamflow of major Ethiopian rivers. Climatic
Change, 143(3), pp.371-383.

Minichil Jemberie, Tesfa Gebrie and Bogale Gebemariam, 2016. Evaluation of land use land
cover change on stream flow: a case study of Dedissa sub basin, Abay basin, south
western Ethiopia. Evaluation, 3(8).

Mishra, V. N., Rai, P. K. (2016): A remote sensing aided multi-layer perceptron Markov
chain analysis for land use and land cover change prediction in Patna district (Bihar),
India. – Arabian Journal of Geosciences 9 (4): 1-18.

Moges Kidane,Alemu Bezie, Nega Kesete and Tefere Tolessa, 2019. The impact of land use
and land cover (LULC) dynamics on soil erosion and sediment yield in
Ethiopia. Heliyon, 5(12), p.e02981.

103
Moss, R.H., Edmonds, J.A., Hibbard, K.A., Manning, M.R., Rose, S.K., Van Vuuren, D.P.,
Carter, T.R., Emori, S., Kainuma, M., Kram, T. and Meehl, G.A., 2010. The next
generation of scenarios for climate change research and
assessment. Nature, 463(7282), pp.747-756.

Mwangi, W.W., Ho, K.W., Tey, B.T. and Chan, E.S., 2016. Effects of environmental factors
on the physical stability of pickering-emulsions stabilized by chitosan particles. Food
Hydrocolloids, 60, pp.543-550.

Natkhin, M., Dietrich, O., Schäfer, M.P. and Lischeid, G., 2015. The effects of climate and
changing land use on the discharge regime of a small catchment in Tanzania. Regional
Environmental Change, 15(7), pp.1269-1280.

Negash Tessema Asfaw Kebede Dame Yadeta, 2020. Modelling the effects of climate change
on streamflow using climate and hydrological models: the case of the Kesem sub-basin
of the Awash River basin, Ethiopia. International Journal of River Basin Management,
pp.1-12.

Ngaina, J., Muthama, N., Mukhala, E. and Maingi, N., 2015. Variability and trends in past,
current and future climate in East Africa.

Nikulin, G., Jones, C., Giorgi, F., Asrar, G., Büchner, M., Cerezo-Mota, R., Christensen, O.B.,
Déqué, M., Fernandez, J., Hänsler, A. and van Meijgaard, E., 2012. Precipitation
climatology in an ensemble of CORDEX-Africa regional climate simulations. Journal
of Climate, 25(18), pp.6057-6078.

Nikulin, G., Lennard, C., Dosio, A., Kjellström, E., Chen, Y., Hänsler, A., Kupiainen, M.,
Laprise, R., Mariotti, L., Maule, C.F. and van Meijgaard, E., 2018. The effects of 1.5
and 2 degrees of global warming on Africa in the CORDEX ensemble. Environmental
Research Letters, 13(6), p.065003.

Ning, T., Li, Z. and Liu, W., 2016. Separating the impacts of climate change and land surface
alteration on runoff reduction in the Jing River catchment of China. Catena, 147,
pp.80-86

104
Obahoundje, S., Ofosu, E.A., Akpoti, K. and Kabo-bah, A.T., 2017. Land use and land cover
changes under climate uncertainty: modelling the impacts on hydropower production
in Western Africa. Hydrology, 4(1), p.2.

Osima, S., Indasi, V.S., Zaroug, M., Hussen Seid., Gudoshava, M., Misiani, H.O., Nimusiima,
A., Anyah, R.O., Otieno, G., Ogwang, B.A. and Jain, S., 2018. Projected climate over
the Greater Horn of Africa under 1.5 C and 2 C global warming. Environmental
Research Letters, 13(6), p.065004.

Parsa, V. A., Yavari, A., Nejadi, A. (2016): Spatio-temporal analysis of land use/land cover
pattern changes in Arasbaran Biosphere Reserve: Iran. – Modeling Earth Systems and
Environment 2 (4): 178.

Neitsch, S.L., Arnold, J.G., Kiniry, J.R. and Williams, J.R., 2011. Soil and water assessment
tool theoretical documentation version 2009. Texas Water Resources Institute.

Pinto, I., Lennard, C., Tadross, M., Hewitson, B., Dosio, A., Nikulin, G., Panitz, H.J. and
Shongwe, M.E., 2016. Evaluation and projections of extreme precipitation over
southern Africa from two CORDEX models. Climatic Change, 135(3-4), pp.655-668.

Pizzigalli, C., Palatella, L., Zampieri, M., Lionello, P., Miglietta, M.M. and Paradisi, P., 2012.
Dynamical and statistical downscaling of precipitation and temperature in a
Mediterranean area. Italian Journal of Agronomy, 7(1), pp.e2-e2.

Reddy, M.A. and Reddy, A., 2008. Textbook of remote sensing and geographical information
systems (pp. 4-4). Hyderabad: BS publications.

Riahi, K., Rao, S., Krey, V., Cho, C., Chirkov, V., Fischer, G., Kindermann, G., Nakicenovic,
N. and Rafaj, P., 2011. RCP 8.5—A scenario of comparatively high greenhouse gas
emissions. Climatic change, 109(1), pp.33-57.

Rummukainen, M., 2010. State‐of‐the‐art with regional climate models. Wiley

Interdisciplinary Reviews: Climate Change, 1(1), pp.82-96.

105
Rwigi, S.K., 2014. Analysis of Potential Impacts of Climate Change and Deforestation on
Surface Water Yields from the Mau Forest Complex Catchments in Kenya (Doctoral
dissertation, University of Nairobi).

Samuel Kassa, 2016. The impact of land use land cover change on watershed hydrology (a
case study of upper awash basin).

Samule Tesfaye, Gebeyehu Taye, Emiru Birhane and S.E van der Zee, 2019. Observed and
model simulated twenty-first century hydro-climatic change of Northern
Ethiopia. Journal of Hydrology: Regional Studies, 22, p.100595.

Shongwe, M.E., Lennard, C., Liebmann, B., Kalognomou, E.A., Ntsangwane, L. and Pinto, I.,
2015. An evaluation of CORDEX regional climate models in simulating precipitation
over Southern Africa. Atmospheric Science Letters, 16(3), pp.199-207.

Sisay Nune, Teshome Soromessa and Demel Teketay 2016. Land use and land cover change
in the Bale mountain eco region of Ethiopia during 1985 to 2015. Land 5:41 . DOI:
10.3390/land5040041

Soni, S., Garg, P.K., Singh, A. and Maurya, A.K., 2015. Assessment of land use land cover
change in Chakrar watershed using geospatial technique. Tropical Plant
Research, 2(2), pp.101-107.

Subedi, P., Subedi, K., Thapa, B. (2013): Application of a hybrid cellular automaton Markov
(CA-Markov) Model in land-use change prediction: a case study of saddle creek
drainage Basin, Florida. – Applied Ecology and Environmental Sciences 1(6): 126-
132.

Subramanya, K., 2013. Hydraulic machines. Tata McGraw-Hill Education.

Kassa Tadele, and Gerd Förch, 2007. Impact of land use/cover change on streamflow: the case
of Hare River Watershed, Ethiopia. In Catchment and lake research, proceedings 2nd
Lake Abaya research symposium (LARS), Arba Minch, Ethiopia.

106
Wakjira Takala, Tamene Adugna and Dawud Tamam, (2016 ) Land use and land cover change
analysis using multi temporal Landsat data in Gilgel Gibe, Omo Gibe basin, Ethiopia.
International journal of science and technology Volume 5 No.7. Center of professional
research publications.

Piani, C., Weedon, G.P., Best, M., Gomes, S.M., Viterbo, P., Hagemann, S. and Haerter, J.O.,
2010. Statistical bias correction of global simulated daily precipitation and temperature
for the application of hydrological models. Journal of hydrology, 395(3-4), pp.199-
215.

Tang, F.F., Xu, H.S. and Xu, Z.X., 2012. Model calibration and uncertainty analysis for runoff
in the Chao River Basin using sequential uncertainty fitting. Procedia Environmental
Sciences, 13, pp.1760-1770.

Tarik Alemayehu, 2017. Modeling the imact of land use/landcover dynamics on stream flow
and sediment yield.

Taylor, K.E., Stouffer, R.J. and Meehl, G.A., 2012. An overview of CMIP5 and the
experiment design. Bulletin of the American meteorological Society, 93(4), pp.485-
498.

Tekalegn Ayele, Nadir Ahmed, L. Ribbe,. and J. Heinrich., 2018. Catchment response to
climate and land use changes in the Upper Blue Nile sub-basins, Ethiopia. Science of
the total environment, 644, pp.193-206.

Temesgen Gashaw, Taffa Tulu, Mekuria Argaw, Abeyou w., Terefe Tolessa and Mengistie
Kindu, 2018. Estimating the impacts of land use/land cover changes on Ecosystem
Service Values: The case of the Andassa watershed in the Upper Blue Nile basin of
Ethiopia. Ecosystem Services, 31, pp.219-228.

Temesgen Gashaw., Amare Bantider., Abraham Mahari. 2014. Population dynamics and land
use/land cover changes in Dera District, Ethiopia. GJBAHS. 3(1):137-140.

Tesfa Gebrie & Bogale Gebremariam, 2015. Impact of Land Use Land Cover Change on
Stream Flow and Sediment Yield : A Case Study of Gilgel Abay Watershed, Lake

107
 Tana Sub-basin, Ethiopia. International Journal of Technology Enhancement and
Emerging Engineering Research, 3, 28–42.

Tsegazeab Dejene, 2014. Tandem Reservoir Operation of Cascade Hydropower Plants Case of
Genale-Dawa River Basin (Doctoral dissertation, Addis Ababa).

Tesfaye Gragn, Asfaw Kebede and Shimelies Berhanu, 2019. Evaluation of Climate Change
Impacts on the Water Resources of Awata River Watershed, Genale Dawa Basin:
Southern Ethiopia.

Teutschbein, C. and Seibert, J., 2012. Bias correction of regional climate model simulations
for hydrological climate-change impact studies: Review and evaluation of different
methods. Journal of hydrology, 456, pp.12-29.

Thomson, A.M., Calvin, K.V., Smith, S.J., Kyle, G.P., Volke, A., Patel, P., Delgado-Arias, S.,
Bond-Lamberty, B., Wise, M.A., Clarke, L.E. and Edmonds, J.A., 2011. RCP4. 5: a
pathway for stabilization of radiative forcing by 2100. Climatic change, 109(1), pp.77-
94.

Trolle, D., Nielsen, A., Andersen, H.E., Thodsen, H., Olesen, J.E., Børgesen, C.D., Refsgaard,
J.C., Sonnenborg, T.O., Karlsson, I.B., Christensen, J.P. and Markager, S., 2019.
Effects of changes in land use and climate on aquatic ecosystems: Coupling of models
and decomposition of uncertainties. Science of the Total Environment, 657, pp.627-
633.

Tufa Negewo and A.K Sarma, 2021. Evaluation of Climate Change-Induced Impact on
Streamflow and Sediment Yield of Genale Watershed, Ethiopia.

Umair, M., Kim, D. and Choi, M., 2019. Impacts of land use/land cover on runoff and energy
budgets in an East Asia ecosystem from remotely sensed data in a community land
model. Science of the Total Environment, 684, pp.641-656.

Van Vuuren, D.P., Lowe, J., Stehfest, E., Gohar, L., Hof, A.F., Hope, C., Warren, R.,
Meinshausen, M. and Plattner, G.K., 2011. How well do integrated assessment models
simulate climate change?. Climatic change, 104(2), pp.255-285.

108
Van Vuuren, D.P., Stehfest, E., den Elzen, M.G., Kram, T., van Vliet, J., Deetman, S., Isaac,
M., Goldewijk, K.K., Hof, A., Beltran, A.M. and Oostenrijk, R., 2011. RCP2. 6:
exploring the possibility to keep global mean temperature increase below 2 C. Climatic
change, 109(1), pp.95-116.

Wakjira Takala, K. Miegel, and Tamene A. Demissie, 2019. Evaluation of the CORDEX
regional climate models performance in simulating climate conditions of two
catchments in Upper Blue Nile Basin. Dynamics of Atmospheres and Oceans, 87,
p.101104.

Wakjira Takala, K. Miegel, and Tamene A. Demissie, 2020. Watershed hydrological response
to combined land use/land cover and climate change in highland Ethiopia: Finchaa
catchment. Water, 12(6), p.1801.\aa

Wang, H., Tetzlaff, D. and Soulsby, C., 2018. Modelling the effects of land cover and climate
change on soil water partitioning in a boreal headwater catchment. Journal of
Hydrology, 558, pp.520-531.

Wayne, J.H., Casper, W.J., Matthews, R.A. and Allen, T.D., 2013. Family-supportive
organization perceptions and organizational commitment: The mediating role of work–
family conflict and enrichment and partner attitudes. Journal of applied
psychology, 98(4), p.606.

Wilby, R.L. and Dawson, C.W., 2013. The statistical downscaling model: insights from one
decade of application. International Journal of Climatology, 33(7), pp.1707-1719.

Wise, M., Calvin, K., Thomson, A., Clarke, L., Bond-Lamberty, B., Sands, R., Smith, S.J.,
Janetos, A. and Edmonds, J., 2009. Implications of limiting CO2 concentrations for
land use and energy. Science, 324(5931), pp.1183-1186.

Woldeamlak Bewket and Solomon Abebe, S., 2013. Land-use and land-cover change and its
environmental implications in a tropical highland watershed, Ethiopia. International
journal of environmental studies, 70(1), pp.126-139.

109
Wu, F., Zhan, J., Yan, H., Shi, C. and Huang, J., 2013. Land cover mapping based on
multisource spatial data mining approach for climate simulation: a case study in the
farming-pastoral ecotone of North China. Advances in Meteorology, 2013.

Wuletaw Abera, Lulseged Tamene, Assefa Abegaz and Dawit Solomon, 2019. Understanding
climate and land surface changes impact on water resources using Budyko framework
and remote sensing data in Ethiopia. Journal of Arid Environments, 167, pp.56-64.

Xu, C.Y., 1999. From GCMs to river flow: a review of downscaling methods and hydrologic
modelling approaches. Progress in physical Geography, 23(2), pp.229-249.

Yang, L., Feng, Q., Yin, Z., Wen, X., Si, J., Li, C. and Deo, R.C., 2017. Identifying separate
impacts of climate and land use/cover change on hydrological processes in upper
stream of Heihe River, Northwest China. Hydrological Processes, 31(5), pp.1100-
1112.

Sahalu ,Atalel, 2014. Analysis of urban land use and land cover changes: a case of study in
Bahir Dar, Ethiopia (Doctoral dissertation).

Yin, J., He, F., Xiong, Y.J. and Qiu, G.Y., 2017. Effects of land use/land cover and climate
changes on surface runoff in a semi-humid and semi-arid transition zone in northwest
China. Hydrology and Earth System Sciences, 21(1), pp.183-196.

Zuo, C., Huang, L., Zhang, M., Chen, Q. and Asundi, A., 2016. Temporal phase unwrapping
algorithms for fringe projection profilometry: A comparative review. Optics and lasers
in engineering, 85, pp.84-103.

Yitea Seneshaw and Van Lanen, 2015. Assessing the impacts of land use-cover change on
hydrology of melka Kuntrie subbasin in Ethiopia, using a conceptual hydrological
model. Hydrology: Current Research, 6(3), p.1.

Zhang, L., Karthikeyan, R., Bai, Z. and Srinivasan, R., 2017. Analysis of streamflow
responses to climate variability and land use change in the Loess Plateau region of
China. Catena, 154, pp.1-11.

110
7. APPENDICES

7.1. Appendix A. Tables

Tables 1: Mean Monthly Flow for Only LULC under Baseline and future Period

Past Land Use land cover Future Land Use land cover
Month LULC1985 LULC 2000 LULC2015 LULC 2035 LULC2055
JAN 18.57 18.63 18.12 18.23 18.48
FEB 12.89 12.42 12.12 12.46 12.05
MAR 20.58 20.16 20.87 21.57 21.92
APR 67.58 67.58 66.59 67.75 68.73
MAY 88.32 89.35 89.64 92.06 93.17
JUN 53.82 54.08 54.16 52.26 51.28
JUL 44.75 43.54 42.64 42.45 41.40
AUG 42.13 43.21 44.84 45.67 45.88
SEP 45.26 45.51 46.98 46.37 46.78
OCT 81.17 82.40 82.87 83.46 84.97
NOV 61.06 62.37 65.19 66.22 67.63
DEC 37.77 37.26 37.11 36.39 35.79
Annual 573.90 576.50 581.14 584.88 588.06

Tables 2: Major soil type of the watershed

No Soil Types Area in (ha) Area in (KM2) Area in %

1 Chromic luvisols 16093.62 160.94 21.95
2 Chromic vertisols 19367.5 193.67 26.42
3 Dystric histosols 784.93 7.85 1.07
4 Eutric cambisols 5004.53 50.05 6.83
5 Eutric fluvisols 11.85 0.12 0.02
6 Eutric nitisols 251.10 2.51 0.34
7 Leptosols 4057.41 40.57 5.53
8 Pellic vertisols 27736.4 277.36 37.84

111
Tables 3: Meteorological data used for SWAT model (1985-2015) for Delo Mena Station

description Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
PCP_MM 25.6
23.81 4 95.21 211.58 210.58 38.64 23.99 37.65 86.54 188.94 87.42 29.59
PCPSTD 3.15 3.33 6.67 9.15 8.66 3.16 2.27 3.42 5.43 8.50 6.86 3.43
PCPSKW 6.53 5.88 3.08 1.57 1.65 3.11 4.04 4.77 3.18 1.70 3.25 5.96
PR_W1 0.06 0.07 0.15 0.41 0.44 0.14 0.14 0.16 0.33 0.33 0.17 0.06
PR_W2 0.47 0.54 0.64 0.73 0.70 0.42 0.42 0.46 0.59 0.75 0.66 0.65
PCPD 3.35 3.71 9.32 18.87 19.42 6.23 6.35 7.58 13.87 18.81 10.84 5.45
TMPMX 31.4
30.46 6 31.27 28.92 27.76 26.91 26.28 28.06 28.61 27.57 28.12 29.27
TMPMN 14.8
13.79 5 15.9 16.78 16.91 16.31 16.06 16.01 16.36 16.19 14.89 13.44
MPSTDMX1 1.69 1.93 2.39 2.68 1.99 2.15 2.25 1.86 1.80 1.81 1.62 1.50
TMPSTDM 1.59 1.61 1.88 1.01 1.25 1.14 1.40 1.44 1.18 1.13 1.71 1.65
RAINHHMX1 12.8
11.91 2 47.61 105.79 105.29 19.32 12.00 18.83 43.27 94.47 43.71 14.80
SOLARAV1 26.1
24.58 8 25.66 23.06 21.83 20.43 19.50 21.18 21.93 21.45 22.70 23.99
DEWPT1 12.0
12.46 6 14.92 18.65 19.22 17.51 16.58 16.85 16.94 17.72 16.54 14.29
WNDAV1 3.06 3.16 3.02 2.57 2.73 3.29 3.17 2.97 2.41 2.52 2.95 3.04
Rira Rainfall Station
PCP_MM 30.0
25.95 5 71.46 158.44 109.15 52.49 109.29 125.82 97.97 115.22 50.94 28.51
PCPSTD 2.46 3.11 4.29 6.52 4.92 3.65 4.40 4.91 4.04 4.93 3.93 2.40
PCPSKW 4.30 2.99 1.66 1.61 4.29 2.01 1.76 1.53 1.59 3.57 3.46 4.30
PR_W1 0.08 0.09 0.20 0.36 0.27 0.23 0.36 0.41 0.40 0.31 0.17 0.10
PR_W2 0.62 0.71 0.77 0.73 0.61 0.77 0.74 0.72 0.75 0.63 0.67 0.62
PCPD 5.71 5.97 12.97 19.23 16.23 11.94 19.77 20.1 18.74 18.29 9.97 8.1
TMPMX 22.21 23.0 23.01 22.01 21.68 21.57 20.85 20.87 21.02 20.69 20.88 21.4
111
 1
TMPMN 8.8 9.27 9.55 10.53 10.68 10.3 9.94 9.99 10.13 10.02 9.37 8.75
MPSTDMX1 1.09 1.21 1.53 1.73 1.72 1.53 1.32 1.48 1.42 1.33 1.27 1.14
TMPSTDM 2.04 1.88 2.62 1.46 1.30 1.55 1.62 1.45 1.46 1.55 1.32 1.89
DEWPT1 7.58 6.78 9.15 12.95 13.37 12.77 12.2 12.16 12.28 12.45 11.13 9.14

112
Tables 4: Mean monthly flow for Climate Change only simulation under RCP4.5 and RCP8.5

Land Use Near future period (2021-2050) and Mid future period (2051-2080)
Month Near –term RCP4.5 Midterm RCP4.5 Near-term RCP8.5 MidtermRCP8.5
JAN 22.99 25.89 24.23 24.92
FEB 11.94 11.61 11.70 10.97
MAR 27.31 26.06 27.23 24.05
APR 77.08 76.77 71.56 71.95
MAY 126.45 124.00 121.00 120.59
JUN 61.91 58.51 59.52 56.72
JUL 35.14 32.61 33.54 32.15
AUG 16.05 16.69 15.25 17.32
SEP 26.92 20.84 26.42 21.47
OCT 95.13 105.59 95.43 106.25
NOV 83.41 86.06 86.09 85.43
DEC 41.13 44.42 42.66 44.57
Annual 625.46 629.06 614.61 616.40

Tables 5: Mean Monthly Flow for Combined Simulation under RCP4.5 and RCP8.5

Land Use 2035 (2021-2050) Land Use 2055(2051-2080)

Month Near-term RCP4.5 Near-term RCP8.5 Midterm RCP4.5 MidtermRCP8.5
JAN 23.05 24.31 26.54 25.53
FEB 11.92 11.69 11.85 11.15
MAR 27.17 27.09 25.06 23.16
APR 77.20 71.73 74.68 70.22
MAY 127.24 121.81 124.23 121.11
JUN 62.91 60.51 62.00 60.28
JUL 35.66 34.05 34.44 33.99
AUG 16.19 15.36 17.15 17.74
SEP 26.79 26.27 20.30 20.82
OCT 95.07 95.33 103.42 103.73
NOV 83.67 86.33 87.14 86.28
DEC 41.48 43.03 46.42 46.44
Annua 628.36 617.52 633.23 620.47
l

112
Tables 6: Projected maximum and minimum Temperature for near and midterm underRCP4.5
and RCP8.5 scenarios

Baseline Max Temp RCP4.5 Max Temp RCP8.5

Month Tp Max Near Mid RCP4.5 Near RCP8.5 MidRCP8.5
RCP4.5
JAN 30.46 31.06 27.67 29.79 31.17
FEB 31.46 30.12 28.36 30.10 31.46
MAR 31.27 28.95 29.24 30.11 31.48
APR 28.92 27.79 30.57 29.71 31.21
MAY 27.76 27.21 31.52 29.43 31.00
JUN 26.91 27.16 32.14 29.22 30.55
JUL 26.28 27.65 31.89 28.97 30.07
AUG 28.06 28.67 31.25 28.95 30.07
SEP 28.61 29.72 30.00 29.41 30.46
OCT 27.57 30.99 28.98 28.96 30.19
NOV 28.12 31.55 27.92 29.19 30.37
DEC 29.27 31.64 27.73 29.36 30.60
Annual 28.72 29.38 29.77 29.43 30.72
Tpmin Min Temp RCP8.5 Min Temp RCP8.5
JAN 14.29 17.93 14.64 18.03 15.35
FEB 14.85 16.95 15.39 16.96 16.33
MAR 15.9 15.77 16.31 15.76 17.47
APR 16.78 14.62 17.67 14.85 18.96
MAY 16.91 14.13 18.58 14.14 20.01
JUN 16.31 14.12 19.12 14.34 20.20
JUL 16.06 14.68 18.80 14.81 19.88
AUG 16.01 15.77 18.10 15.83 18.94
SEP 16.36 16.85 16.83 17.08 17.71
OCT 16.19 18.12 15.77 18.10 16.49
NOV 15.29 18.62 14.73 18.67 15.47
DEC 15.44 18.63 14.66 18.49 15.23
Annual 15.87 16.35 16.72 16.42 17.67

113
Tables 7: Projected mean monthly and % of rainfall under RCP4.5 and RCP8.5.

RCP4.5 RCP8.5

NRF % MRF % NRF % MRF %

Month A
JAN 23.81 29.10 22.20 29.35 23.28 28.28 18.79 29.42 23.55
FEB 25.64 27.96 9.05 25.15 -1.90 26.76 4.38 25.99 1.38
MAR 95.21 96.84 1.71 91.82 -3.56 95.59 0.40 91.17 -4.24
APR 211.58 197.65 -6.58 198.94 -5.97 191.8 -9.38 199.80 -5.57
MAY 210.58 204.30 -2.98 210.50 -0.04 205.3 -2.50 210.12 -0.22
JUN 38.64 32.68 -15.43 32.51 -15.87 33.42 -13.52 32.99 -14.62
JUL 23.99 26.30 9.64 27.37 14.10 25.95 8.16 27.58 14.95
AUG 37.65 34.49 -8.38 40.56 7.74 31.95 -15.14 41.78 10.98
SEP 86.54 100.97 16.67 86.72 0.21 98.68 14.03 87.92 1.60
OCT 188.94 190.07 0.60 209.22 10.74 192.6 1.94 208.22 10.20
NOV 87.42 93.77 7.27 92.43 5.73 96.35 10.22 93.50 6.96
DEC 29.59 26.83 -9.32 28.27 -4.46 28.10 -5.04 28.65 -3.19
Nate: A= Represent observed rainfall, NRF = Near-term Rainfall, MRF= midterm rainfall
and % = percentage change of future rainfall from baseline period.

Figures 1: Thiessen polygon of the station

114